mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-16 03:31:44 +01:00
perf(ci): shorten the general server and chat test lanes (#21393)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work > - Every pull request runs the `PR` CI workflow before it can merge > - The `General tests (server (N/12))` and `General tests (chat (N/3))` jobs finish last, so they set the wall-clock time of CI > - Before this change, server shards took 9.5 to 11.7 minutes and chat shards took 6.4 to 9.6 minutes; the target is 7 minutes or less > - Measurement shows that the chat lane was unbalanced and that the server lane is limited by CPU on the 4-vCPU runners > - This pull request balances the chat lane by recorded test time, runs three chat partitions per job, and takes general server weights from PR runs > - The benefit is a chat lane under 7 minutes and measured evidence for the shard count that the server lane needs ## Linked Issues or Issue Description **What existing behavior does this improve?** The PR CI general server and chat test jobs. They are the slowest jobs of every pull request. **Subsystem affected** CI test scheduling: `scripts/run-vitest-stable.mjs`, `scripts/test-line-shard.mjs` and `scripts/ci/refresh-test-durations.mjs`. **Current behavior** Baseline: green `PR` runs [43437558012](https://github.com/paperclipai/paperclip/actions/runs/43437558012) and [43441184316](https://github.com/paperclipai/paperclip/actions/runs/43441184316), all on m7i-flex.xlarge (4 vCPU, 16 GB). - Server shards: 9.5 to 11.7 job minutes, 502 to 654 s test step, effective file concurrency 2.5 to 2.9 of 3 workers. Vitest reports `tests 61-68%, import 25-31%, transform 6-7%, setup 1%`. Each shard sums 1,410 to 1,765 s of file time. The longest file is 303 s (`deferred-execution-receipt-work.test.ts`), so the floor is total work divided by workers, not one file. - Chat shards: 9.3 to 9.6, 7.6 to 7.8 and 6.4 to 6.5 job minutes. Each job ran two partitions of `chat-channels.integration.test.ts`. The partitions were balanced by collected-case count, but one `it.each` line can run many cases. One of six partitions recorded 532 s of test time against a 377 s mean. **Proposed behavior** - Chat partitions keep each source line whole and balance by recorded time per collected entry. - Each chat job runs three partitions in place of two. - General server weights come from the server jobs of merged PR runs, which run files with the same concurrency as the lane they place. **Reason and benefit** The slowest job sets the CI wall-clock time for every pull request. The chat lane goes from a 9.3-9.6 minute maximum to 6.5-6.8 minutes. ## What Changed - `scripts/test-line-shard.mjs`: `partitionTestLines` takes optional recorded durations. With durations, it balances source lines by recorded time; entries without a record take the median. Without durations, the behavior is unchanged. - `scripts/chat-shard-durations.json` (new): recorded time per collected chat entry, from the CI profile records of PR run [43444854171](https://github.com/paperclipai/paperclip/actions/runs/43444854171). 926 of 928 collected entries are measured; the other two are the bogus collection entries that the launcher already drops. - `scripts/ci/refresh-chat-shard-durations.mjs` (new): builds that manifest from downloaded `ci-profile-*` artifacts. Each executed case maps to the collected `it`/`it.each` entry that registers it. A case without an entry fails the refresh. - `scripts/run-vitest-stable.mjs`: a sharded chat job runs three partitions in place of two, and the chat launcher passes the recorded durations. A chat job collects the suite once with `vitest list` in place of once per partition, and the per-partition filter checks run at the same time. Before, each partition ran two blocking collections of the 86,000-line file one after another. The server worker count stays at three. - `scripts/ci/refresh-test-durations.mjs` and `scripts/ci/pr-run-durations.mjs`: the weekly refresh reads general server weights from the two newest merged-PR runs. Each run must have the full server shard count from `.github/ci-task-matrix.json` and cover at least 90% of the lane. Serialized weights still come from master Cloud readiness runs. - `scripts/general-server-shard-durations.json`: refreshed with the new source (PR runs 43439145752 and 43439279537). - Tests: `scripts/__tests__/run-vitest-stable-shard.test.mjs`, `scripts/ci/test-selection-pr-timings.test.mjs` and `server/src/__tests__/vitest-chat-shards.test.ts` cover the weighted partition, the entry mapping, the manifest format, the PR-run collection rules and the three-partition chat launcher. - `doc/DEVELOPING.md`: describes the three chat partitions, the chat weights and how to refresh them, and the new source of the server weights. ## Verification Experiments ran in this PR's own CI with a temporary resource sampler (removed before review). | Setting (run) | Server step vs. same shard's baseline | CPU busy | Max memory used | Result | |---|---|---|---|---| | 3 workers (43444854171, shards 1-4) | +8% (profiling overhead) | 62-73% | 6.8 GB | 1 flaky retry | | 4 workers (43444854171, shards 5-8) | -4% | 82-86% | 8.2 GB | clean | | 5 workers (43444854171, shards 9-12) | -10% | 70-93% | 9.5 GB | 1 hard failure: `decision-models.test.ts` timed out at 15 s twice | | 4 workers (43450469416) | mixed m5a/m5n/m6i runners | 78-89% | 7.7 GB | 1 hard failure: `workspace-recovery-warm-handoff.test.ts` timed out at 30 s on m5a | The server lane is CPU bound: at four workers the CPU is 82-89% busy and wall time improves by only about 10%. Five workers fail tests. Memory is not the limit (no swap, no memory pressure, at least 6 GB free). Node's compile cache and Postgres `fsync=off` showed no attributable gain (import share stayed 29-34%; `fsync=off` cut iowait from 2-4% to 0.2-0.7% of CPU time), so neither is in this PR. Final configuration (server: 3 workers with refreshed weights; chat: 3 weighted partitions per job), job minutes / test-step seconds / effective concurrency: | Shard | Baseline 43437558012 | Baseline 43441184316 | A: 43465156503 attempt 1 | B: attempt 2 | C: attempt 3 | |---|---|---|---|---|---| | server 1/12 | 11.0 / 609 / 2.81 | 10.2 / 562 / 2.82 | 10.1 / 535 / 2.88 | 11.6 / 615 / 2.86 | 11.3 / 600 / 2.86 | | server 2/12 | 10.2 / 567 / 2.71 | 10.9 / 607 / 2.79 | 9.7 / 510 / 2.85 | 8.9 / 466 / 2.86 | 11.9 / 635 / 2.83 | | server 3/12 | 10.9 / 609 / 2.88 | 10.9 / 601 / 2.86 | 9.4 / 498 / 2.81 | 11.7 / 623 / 2.80 | 20.3 / 1109 / 0.00 | | server 4/12 | 11.0 / 613 / 2.77 | 11.2 / 622 / 2.84 | 11.3 / 602 / 2.84 | 9.3 / 487 / 2.83 | 12.5 / 676 / 2.84 | | server 5/12 | 10.1 / 555 / 2.83 | 9.9 / 544 / 2.84 | 10.2 / 542 / 2.85 | 11.6 / 618 / 2.82 | 11.9 / 640 / 2.84 | | server 6/12 | 9.9 / 551 / 2.85 | 10.4 / 579 / 2.85 | 9.8 / 516 / 2.69 | 9.1 / 480 / 2.68 | 9.2 / 480 / 2.67 | | server 7/12 | 10.8 / 596 / 2.84 | 10.1 / 558 / 2.86 | 10.8 / 575 / 2.80 | 11.8 / 631 / 2.80 | 20.8 / 1129 / 2.69 | | server 8/12 | 11.7 / 654 / 2.62 | 11.2 / 624 / 2.56 | 9.8 / 513 / 2.83 | 11.2 / 589 / 2.82 | 19.4 / 1059 / 2.40 | | server 9/12 | 9.7 / 533 / 2.85 | 9.7 / 533 / 2.84 | 11.5 / 615 / 2.68 | 9.8 / 515 / 2.66 | 13.0 / 703 / 2.68 | | server 10/12 | 9.8 / 540 / 2.79 | 9.5 / 502 / 2.81 | 10.6 / 561 / 2.84 | 11.8 / 625 / 2.83 | 9.3 / 488 / 2.83 | | server 11/12 | 10.9 / 605 / 2.83 | 11.3 / 627 / 2.75 | 10.3 / 545 / 2.75 | 9.5 / 498 / 2.72 | 20.3 / 1107 / 0.00 | | server 12/12 | 10.6 / 592 / 2.50 | 11.7 / 645 / 2.53 | 9.6 / 503 / 2.81 | 8.9 / 468 / 2.82 | 19.7 / 1065 / 2.70 | | chat 1/3 | 9.6 / 528 / 1.56 | 9.3 / 511 / 1.58 | 6.6 / 310 / 2.69 | 5.7 / 272 / 2.65 | 6.8 / 331 / 2.69 | | chat 2/3 | 7.8 / 418 / 1.74 | 7.6 / 408 / 1.75 | 6.3 / 305 / 2.57 | 6.5 / 311 / 2.58 | 5.5 / 260 / 2.58 | | chat 3/3 | 6.5 / 337 / 1.74 | 6.4 / 333 / 1.73 | 6.0 / 285 / 2.65 | 5.4 / 258 / 2.68 | 5.8 / 262 / 2.69 | Summary: chat maximum 9.3-9.6 min before, 6.6 / 6.5 / 6.8 min in A / B / C. Server maximum 11.7 min before, 11.5 / 11.8 min in A / B; the server worker setting is unchanged, so its run-to-run spread is mostly runner type (m8i.xlarge shards took 8.9-9.8 min, m6a.xlarge 11.9-13.0 min). Runners: A all m7i.xlarge; B m8i/m6i; C m6a/m6in/m8i/m5a. In C, the five server shards on m5a.xlarge summed 2,545-3,037 s of file time against about 1,700 s elsewhere; three reached the 20-minute job limit and two failed tests on 15 s timeouts. The server lane settings in this PR are the same as master's, so this is a fleet finding: m5a.xlarge cannot run a server shard inside the job limit (follow-up for the fleet configuration, outside this PR). Flakiness of the final setting: the nine chat jobs in A, B and C had no failure and no retried test. Server shards in A and B had no failure and no retried test. Before the last commit (single chat collection), three runs of 43456289623 had chat at 6.7-7.5 min with no chat retries; server retries there (2 in one run, 1 in another) were on the unchanged three-worker server setting. Local checks (correctness only): - `node --test scripts/__tests__/run-vitest-stable-shard.test.mjs scripts/ci/test-selection-pr-timings.test.mjs scripts/ci/test-selection-timings.test.mjs`: 68 passed. - `pnpm exec vitest run --project @paperclipai/server server/src/__tests__/vitest-chat-shards.test.ts`: passed. - `pnpm verify` exits 2 (global CI input changed), so the required CI checks supply the broad coverage. ## Risks - Three chat partitions per job use more memory: the peak was 11.4 GB of 16 GB, with at least 4.2 GB free and no memory pressure. CPU was 55-63% busy. - Stale chat weights only change balance, never which tests run. Shard membership is still checked with `vitest list` before each run. - The weekly refresh now downloads general server logs from merged-PR runs. Runs that selected a subset of files are skipped before download because their shard count is smaller. - The server lane is still above 7 minutes. See the remaining floor below. **Remaining floor and shard-count recommendation (for a follow-up under `.github/`)** A server shard sums about 1,600 s of file time, and 3 workers on 4 vCPUs give at most about 2.8 effective concurrency, so the test step is about 560-640 s. Job overhead is about 50 s. A job of 7 minutes needs a test step of about 360 s, so about 1,000 s of file time per shard. The current 19,200 s of general server file time therefore needs about 19-20 shards of 4 vCPUs at 3 workers. More workers on the same runners do not help, because the CPU is already the limit. ## Model Used - Claude Opus 5.5 (`claude-opus-5-5`), Anthropic, with tool use (shell, GitHub CLI) in Claude Code. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge 🤖 Generated with [Claude Code](https://claude.com/claude-code) --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
0199c963e2
commit
0d832a02d3
11 files changed
+3042
-1769
No files matched your search
+23
-4
@@ -332,8 +332,16 @@ repository's complete history.
|
||||
|
||||
Server test costs are recorded in CI with tests, hooks, collection, setup and
|
||||
environment preparation included. The weekly **Refresh CI test durations**
|
||||
workflow reads two successful protected-master Cloud readiness runs and opens a
|
||||
timing-only PR. It never executes code from logs, changes test selection rules,
|
||||
workflow opens a timing-only PR. It reads the general server weights from the
|
||||
`General tests (server)` jobs of the two newest successful `PR` runs of merged
|
||||
PRs, pinned to each PR's final head. Each run must have the full shard count in
|
||||
`.github/ci-task-matrix.json` and cover at least 90% of the lane, so a run that
|
||||
selected a few files does not count. These jobs run files three at a time, and
|
||||
their contention changes relative file costs. Measured against two other PR
|
||||
runs, Cloud readiness weights placed the slowest of 12 shards at 1,748-1,770 s of
|
||||
summed file time, against a 1,595 s mean; PR-run weights placed it at 1,684-1,702 s. The
|
||||
serialized weights still come from two successful protected-master Cloud
|
||||
readiness runs. It never executes code from logs, changes test selection rules,
|
||||
or grants AWS access. GitHub may require a maintainer to approve the bot PR's
|
||||
workflow runs; the refresh never merges itself. An outstanding refresh PR pauses
|
||||
new refreshes. Both timing manifests include source runs and measurement coverage.
|
||||
@@ -439,8 +447,19 @@ recovery behavior for the same reason. Unsharded local tests and route/authz
|
||||
lanes retain their existing file concurrency. Maintenance server cases overlap
|
||||
two owned server/database fixtures within their file; an explicit shared
|
||||
maintenance database keeps them serial. Sharded chat jobs run
|
||||
two independent partitions, each with its own embedded PostgreSQL, home and
|
||||
config; an explicit shared test database retains one partition per job.
|
||||
three independent partitions, each with its own embedded PostgreSQL, home and
|
||||
config; an explicit shared test database retains one partition per job. The
|
||||
partitions keep each source line whole and balance by the recorded time in
|
||||
`scripts/chat-shard-durations.json`; entries without a record take the median.
|
||||
Collected-case counts were a poor proxy because one `it.each` line can run many
|
||||
cases: one of six count-balanced partitions recorded 532 s of test time against
|
||||
a 377 s mean. To refresh the weights, add the `profile-ci` label to a PR, download
|
||||
the `ci-profile-*` artifacts of its three chat jobs into one directory, and run
|
||||
`node scripts/ci/refresh-chat-shard-durations.mjs --source=<run URL> <directory>`.
|
||||
Stale weights change only the balance, never which tests run. A chat job
|
||||
collects the suite once, then validates its three partitions' filters at the
|
||||
same time; each partition still checks that its filters select exactly its
|
||||
assigned tests.
|
||||
The Runner transport lease-expiry and cold-admission regressions live in their
|
||||
own file so those real-clock waits overlap other files in the existing pool.
|
||||
|
||||
|
||||
@@ -14,6 +14,7 @@ import {
|
||||
|
||||
import { assertManifestSource } from "../ci/pr-run-durations.mjs";
|
||||
import { assertSelectedTests, partitionTestLines } from "../test-line-shard.mjs";
|
||||
import { chatManifestData, entryDurations } from "../ci/refresh-chat-shard-durations.mjs";
|
||||
|
||||
const repoRoot = path.resolve(path.dirname(fileURLToPath(import.meta.url)), "..", "..");
|
||||
const script = path.join(repoRoot, "scripts", "run-vitest-stable.mjs");
|
||||
@@ -606,6 +607,47 @@ test("test-line shards cover nested and parameterized cases exactly once without
|
||||
assert.deepEqual(partitionTestLines([...cases].reverse(), 3, lineShardFile).map((shard) => shard.lines), shards.map((shard) => shard.lines));
|
||||
});
|
||||
|
||||
test("recorded durations balance chat line shards by time without splitting a source line", () => {
|
||||
const cases = [caseAt(10, "slow"), caseAt(20, "each %s"), caseAt(20, "each %s"), caseAt(30, "a"), caseAt(40, "b"),
|
||||
caseAt(50, "c"), caseAt(60, "unrecorded")];
|
||||
const durations = { slow: 9000, "each %s": 1000, a: 3000, b: 3000, c: 2000 };
|
||||
const shards = partitionTestLines(cases, 2, lineShardFile, durations);
|
||||
assert.deepEqual(shards.map((shard) => shard.weight), [11000, 11000]);
|
||||
assert.deepEqual(shards.map((shard) => shard.lines), [[10, 20], [30, 40, 50, 60]]);
|
||||
assert.deepEqual(shards.flatMap((shard) => shard.tests).sort((a, b) => a.location.line - b.location.line), cases);
|
||||
assert.deepEqual(partitionTestLines([...cases].reverse(), 2, lineShardFile, durations).map((shard) => shard.lines),
|
||||
shards.map((shard) => shard.lines));
|
||||
assert.deepEqual(partitionTestLines(cases, 2, lineShardFile).map((shard) => shard.tests.length), [4, 3]);
|
||||
});
|
||||
|
||||
test("chat durations attribute every executed case to the collected entry that registers it", () => {
|
||||
const entries = ["sends %s", "suite > sends %s twice", "plain", "keyed $provider ($mode)"];
|
||||
const { totals, unmatched } = entryDurations(entries, [
|
||||
{ name: "chat channel control-plane integration > sends a", duration: 5 },
|
||||
{ name: "chat channel control-plane integration > suite > sends b twice", duration: 7 },
|
||||
{ name: "chat channel control-plane integration > plain", duration: 11 },
|
||||
{ name: "chat channel control-plane integration > keyed slack (fast)", duration: 13 },
|
||||
{ name: "chat channel control-plane integration > keyed github (slow)", duration: 17 },
|
||||
{ name: "chat channel control-plane integration > unknown", duration: 19 },
|
||||
]);
|
||||
assert.deepEqual(Object.fromEntries(totals), { "sends %s": 5, "suite > sends %s twice": 7, plain: 11, "keyed $provider ($mode)": 30 });
|
||||
assert.deepEqual(unmatched, ["chat channel control-plane integration > unknown"]);
|
||||
const data = chatManifestData(["plain", "other"], [[{ name: "x > plain", duration: 4 }], [{ name: "x > plain", duration: 9 }],
|
||||
[{ name: "x > plain", duration: 6 }]], []);
|
||||
assert.deepEqual(data.durations, { plain: 6 });
|
||||
assert.deepEqual(data.coverage, { measured: 1, entries: 2 });
|
||||
assert.throws(() => chatManifestData(["plain"], [[{ name: "x > missing", duration: 1 }]], []));
|
||||
});
|
||||
|
||||
test("the checked-in chat duration manifest is well formed", () => {
|
||||
const manifest = JSON.parse(readFileSync(path.join(repoRoot, "scripts", "chat-shard-durations.json"), "utf8"));
|
||||
assert.equal(manifest.unit, "ms");
|
||||
const entries = Object.entries(manifest.durations);
|
||||
assert.ok(entries.length > 0);
|
||||
assert.ok(entries.every(([name, ms]) => name.length > 0 && Number.isSafeInteger(ms) && ms > 0));
|
||||
assert.deepEqual(loadShardDurations(path.join(repoRoot, "scripts", "chat-shard-durations.json")), manifest.durations);
|
||||
});
|
||||
|
||||
test("line-shard collection rejects empty, foreign, or unlocated tests and invalid shard counts", () => {
|
||||
const good = caseAt(10, "valid");
|
||||
for (const input of [[], null, [{ ...good, file: "/another.test.ts" }], [{ ...good, projectName: "wrong" }],
|
||||
@@ -632,10 +674,10 @@ test("parallel chat jobs form a complete disjoint partition", () => {
|
||||
dryRunJson(["--mode", "general", "--group", "general-chat", "--shard-index", String(index), "--shard-count", String(count)],
|
||||
{ PAPERCLIP_TEST_DATABASE_URL: "" }).chatVitestShards,
|
||||
);
|
||||
assert.ok(shards.every((workers) => workers.length === 2));
|
||||
assert.ok(shards.every((workers) => workers.length === 3));
|
||||
assert.deepEqual(shards.flat().map(({ index }) => index).sort((a, b) => a - b),
|
||||
Array.from({ length: count * 2 }, (_, index) => index));
|
||||
assert.ok(shards.flat().every((shard) => shard.count === count * 2));
|
||||
Array.from({ length: count * 3 }, (_, index) => index));
|
||||
assert.ok(shards.flat().every((shard) => shard.count === count * 3));
|
||||
}
|
||||
});
|
||||
|
||||
@@ -784,8 +826,8 @@ test("a failed chat child fails the job after its sibling finishes", { skip: pro
|
||||
const directory = mkdtempSync(path.join(os.tmpdir(), "chat-launch-proof-"));
|
||||
const file = path.join(repoRoot, "server/src/__tests__/chat-channels.integration.test.ts");
|
||||
const lines = readFileSync(file, "utf8").split("\n")
|
||||
.flatMap((line, index) => /^\s*it\(/.test(line) ? [index + 1] : []).slice(0, 6);
|
||||
assert.equal(lines.length, 6);
|
||||
.flatMap((line, index) => /^\s*it\(/.test(line) ? [index + 1] : []).slice(0, 9);
|
||||
assert.equal(lines.length, 9);
|
||||
const tests = lines.map((line) => ({ file, projectName: "@paperclipai/server",
|
||||
name: `launcher fixture ${line}`, location: { line } }));
|
||||
const receipt = path.join(directory, "sibling-finished");
|
||||
|
||||
@@ -0,0 +1,939 @@
|
||||
{
|
||||
"$comment": "Generated by node scripts/ci/refresh-chat-shard-durations.mjs from the CI profile test records of the PR chat jobs (profile-ci label). Per collected test entry, keyed by the name that vitest list reports: the summed duration of every case that entry runs, so an it.each line carries all its cases. Median across the given runs. scripts/test-line-shard.mjs uses these weights only to place source lines on chat partitions; they never change which tests run. Missing entries take the median weight.",
|
||||
"unit": "ms",
|
||||
"sources": [
|
||||
"https://github.com/paperclipai/paperclip/actions/runs/43444854171"
|
||||
],
|
||||
"coverage": {
|
||||
"measured": 926,
|
||||
"entries": 928
|
||||
},
|
||||
"durations": {
|
||||
"$label": 2632,
|
||||
"accepts a manually created GitHub App when /app lists only selectable webhook events": 376,
|
||||
"accepts only signed GitHub setup pings before App credentials exist": 174,
|
||||
"accepts the Slack invitation mention exactly once (%s)": 1081,
|
||||
"acknowledges a denied Slack action before its durable notice completes and deduplicates redelivery": 789,
|
||||
"acknowledges a provider question when the board resolves after callback validation": 760,
|
||||
"acknowledges a signed URL mention through the real Slack SDK without waiting for an unfurl": 378,
|
||||
"acknowledges a Slack mention while %s is blocked": 1508,
|
||||
"acknowledges a Slack retry while one durable worker owns the unrecorded command root": 428,
|
||||
"acknowledges denied Telegram actions after one durable, payload-free notice": 278,
|
||||
"acknowledges durable Slack ingress promptly and never replays paused traffic on resume": 961,
|
||||
"acknowledges GitHub lifecycle callbacks without changing a paused endpoint": 384,
|
||||
"acknowledges Telegram control $command without a processing receipt (bound=$bound denied=$denied)": 2880,
|
||||
"acknowledges Telegram edits while verifying and when processing is suspended": 361,
|
||||
"activates Slack only after provider verification and a real test message": 760,
|
||||
"admits a separate connected-conversation Retry only after committed original-controller containment: %s": 7416,
|
||||
"admits concurrent and retried Slack slash commands only once": 527,
|
||||
"admits current GitHub task Retry without replaying obsolete input: %s": 5540,
|
||||
"admits only an addressed raw GitHub comment as the first setup repository": 690,
|
||||
"admits ordinary chat task Retry with a missing checkpoint only after containment: %s": 3063,
|
||||
"admits the exact chat source for ordinary %s failure": 1312,
|
||||
"advances failed retry notices past denied history with microsecond timestamps (%s)": 1551,
|
||||
"advances native progress past suppressed candidates with microsecond timestamps": 1012,
|
||||
"advances pre-run FIFO selection past a suppressed page with microsecond timestamps (%s)": 2286,
|
||||
"allows a new exact failed chat retry after an ordinary retry fails, without rearming its ancestors": 518,
|
||||
"allows only one concurrent live endpoint to claim a Discord bot across guilds": 251,
|
||||
"allows Slack thread replies without mentions while exempting DMs (DM=%s)": 1850,
|
||||
"applies Teams installation and Telegram membership resource lifecycle": 972,
|
||||
"applies the direct-message toggle to Slack slash commands": 315,
|
||||
"applies the Slack Board resume once during %s dispatch": 1465,
|
||||
"applies the Teams direct-message reach toggle and starts a new generation after task completion": 564,
|
||||
"atomically admits only one first GitHub repository under concurrent root mentions": 740,
|
||||
"atomically rejects a Slack DM when reach is revoked after initial admission": 324,
|
||||
"atomically rejects a sponsored Telegram guest after the sponsor is suspended": 224,
|
||||
"atomically rejects a Telegram DM when reach is revoked at the final task-mutation boundary": 214,
|
||||
"atomically rejects Slack task mutation after a linked identity is revoked or loses write membership": 492,
|
||||
"atomically rejects Teams messages when DM, group, or channel reach is revoked at the task-mutation boundary": 815,
|
||||
"audits Microsoft Teams reactions idempotently without treating them as task instructions": 683,
|
||||
"audits reactions on linked messages without treating them as task instructions": 770,
|
||||
"audits repeated Discord reaction cycles while deduplicating an exact Gateway replay": 732,
|
||||
"audits Telegram reactions on completed DM generations idempotently without comments or runs": 792,
|
||||
"authorizes protected channel workspace defaults for the endpoint's assigned agent": 501,
|
||||
"automatic Slack registration > acknowledges verification before sending one DM across concurrent callbacks and restart recovery": 638,
|
||||
"automatic Slack registration > adds events only after saving signing credentials and answers Slack's update-time challenge before OAuth": 722,
|
||||
"automatic Slack registration > claims callbacks once, keeps denied installation resumable, and does not replay ambiguous exchange": 726,
|
||||
"automatic Slack registration > creates once under concurrent submissions and resumes after restart without exposing credentials": 295,
|
||||
"automatic Slack registration > does not advance avatar state after management access is revoked during upload": 159,
|
||||
"automatic Slack registration > does not bind runtime credentials when management access is revoked during provider checks": 323,
|
||||
"automatic Slack registration > does not complete setup from %s event evidence": 7684,
|
||||
"automatic Slack registration > does not link or welcome an installer after company membership is revoked during lookup": 435,
|
||||
"automatic Slack registration > does not repeat uncertain creation until the operator explicitly confirms no app exists": 240,
|
||||
"automatic Slack registration > does not replay a verification DM after %s": 1457,
|
||||
"automatic Slack registration > does not replay a welcome dispatch interrupted by restart": 683,
|
||||
"automatic Slack registration > does not send a pending verification DM after %s": 2415,
|
||||
"automatic Slack registration > does not send a pending welcome to a subsequently revoked account link": 716,
|
||||
"automatic Slack registration > does not verify or notify for a %s challenge": 3132,
|
||||
"automatic Slack registration > keeps installation usable after %s without leaking provider echoes or creating another app": 1451,
|
||||
"automatic Slack registration > keeps setup connected after %s and never retries an optional DM": 2017,
|
||||
"automatic Slack registration > keeps the saved app installable after local avatar rendering fails": 166,
|
||||
"automatic Slack registration > links the OAuth installer, sends one welcome DM, and still requires signed URL verification": 1363,
|
||||
"automatic Slack registration > managed Slack > atomically consumes manager state and rejects wrong actors and expired attempts": 163,
|
||||
"automatic Slack registration > managed Slack > claims concurrent manager callbacks once": 287,
|
||||
"automatic Slack registration > managed Slack > clears a manager reconnect error after renewing the bound grant": 1086,
|
||||
"automatic Slack registration > managed Slack > creates and installs once, uploads the avatar and links only the manager grant owner": 1398,
|
||||
"automatic Slack registration > managed Slack > does not accept another company's grant or keep authorizations after removal": 265,
|
||||
"automatic Slack registration > managed Slack > does not activate a bot when the manager grant is revoked during child OAuth": 639,
|
||||
"automatic Slack registration > managed Slack > does not restore a revoked grant when an in-flight refresh returns": 257,
|
||||
"automatic Slack registration > managed Slack > installs a second bot using the same grant without new manager consent": 1594,
|
||||
"automatic Slack registration > managed Slack > keeps configured Cloud drafts managed when the broker profile disappears": 198,
|
||||
"automatic Slack registration > managed Slack > keeps open-source setup independent of the broker": 202,
|
||||
"automatic Slack registration > managed Slack > never recreates an uncertain app": 343,
|
||||
"automatic Slack registration > managed Slack > preserves configured-app reconnect intent when managed installation fails": 1569,
|
||||
"automatic Slack registration > managed Slack > preserves declined manager authorization as a safe recoverable outcome": 215,
|
||||
"automatic Slack registration > managed Slack > preserves provider rate-limit backoff across subsequent attempts": 492,
|
||||
"automatic Slack registration > managed Slack > reconnects a rejected manager token without losing pending manifest configuration": 1044,
|
||||
"automatic Slack registration > managed Slack > reinstalls the configured app on an explicit %s reconnect": 4077,
|
||||
"automatic Slack registration > managed Slack > reinstalls the saved app after missing bot scopes without creating another app": 1053,
|
||||
"automatic Slack registration > managed Slack > rejects a bot token belonging to a different Slack app before activation": 499,
|
||||
"automatic Slack registration > managed Slack > rejects manager identity changes and both callback-origin and ingress-origin drift": 182,
|
||||
"automatic Slack registration > managed Slack > rejects missing tool scopes before activating a managed bot": 432,
|
||||
"automatic Slack registration > managed Slack > rejects missing, malformed, foreign and other-user manager grants before provisioning": 232,
|
||||
"automatic Slack registration > managed Slack > resumes saved installation credentials after inventory failure without creating or reinstalling": 996,
|
||||
"automatic Slack registration > managed Slack > retains reinstall intent after a provider failure instead of retrying rejected credentials": 977,
|
||||
"automatic Slack registration > managed Slack > reuses the same personal workspace grant for another draft without another authorization": 210,
|
||||
"automatic Slack registration > managed Slack > revocation invalidates pending authorization and prevents future provisioning": 229,
|
||||
"automatic Slack registration > managed Slack > routes the broker claim_id callback back to the saved wizard without exposing the handoff": 294,
|
||||
"automatic Slack registration > managed Slack > serializes refreshes across endpoints and never replays an ambiguous refresh": 234,
|
||||
"automatic Slack registration > managed Slack > uses bound OAuth for approval fallback and rejects a different installer": 555,
|
||||
"automatic Slack registration > managed Slack > uses the held transaction for grant queries without acquiring another pool connection": 405,
|
||||
"automatic Slack registration > maps invalid token and validation echoes without persisting or exposing the supplied token": 253,
|
||||
"automatic Slack registration > preserves a saved app when its creation activity fails without exposing the failure": 189,
|
||||
"automatic Slack registration > preserves an existing %s account link without a welcome DM": 1056,
|
||||
"automatic Slack registration > recovers %s on the same saved app after restart without exposing provider echoes": 3290,
|
||||
"automatic Slack registration > reinstalls a revoked bot using the same app and preserves the linked account": 1033,
|
||||
"automatic Slack registration > rejects a missing or invalid OAuth installer (%s)": 504,
|
||||
"automatic Slack registration > rejects an authorization started before the bot was revoked": 709,
|
||||
"automatic Slack registration > rejects an installer lookup %s mismatch without linking": 2093,
|
||||
"automatic Slack registration > rejects an OAuth %s mismatch without binding runtime credentials": 1061,
|
||||
"automatic Slack registration > rejects changing an established %s during reauthorization": 1306,
|
||||
"automatic Slack registration > rejects expiry, wrong actor/session/company, permission revocation, and origin drift before exchange": 287,
|
||||
"automatic Slack registration > removes registration secrets and prevents late callbacks after %s removal": 590,
|
||||
"automatic Slack registration > respects disabled direct messages when provisioning the welcome": 512,
|
||||
"automatic Slack registration > resumes an interrupted avatar upload without repeating creation or requesting another token": 232,
|
||||
"automatic Slack registration > retains the token and installer after lookup failure and resumes without exchanging a code": 953,
|
||||
"automatic Slack registration > returns expired authorization to the saved wizard without exchanging its code": 220,
|
||||
"automatic Slack registration > reuses the saved installer and sends no further DM on reauthorization or restart": 1072,
|
||||
"automatic Slack registration > saves advanced app details with the initial draft and rejects them for other providers": 145,
|
||||
"automatic Slack registration > shows interrupted creation as uncertain after restart and supports manual recovery on the same draft": 129,
|
||||
"automatic Slack registration > Slack test messages > does not downgrade confirmed delivery when activity logging fails": 1050,
|
||||
"automatic Slack registration > Slack test messages > does not restore revoked access or take another user's identity": 841,
|
||||
"automatic Slack registration > Slack test messages > loads another workspace page instead of truncating account discovery": 950,
|
||||
"automatic Slack registration > Slack test messages > redacts provider failures and never retries an ambiguous send": 1229,
|
||||
"automatic Slack registration > Slack test messages > rejects wrong-workspace users, disabled DMs, and membership revocation": 898,
|
||||
"automatic Slack registration > Slack test messages > requires a private Slack challenge for an unknown account before linking and sending the test": 1019,
|
||||
"automatic Slack registration > Slack test messages > sends to the current linked user and reuses a durable result on duplicate requests": 1142,
|
||||
"automatic Slack registration > Slack test messages > settles a failed Slack account-link intent without committing identity or sending a test message": 1009,
|
||||
"automatic Slack registration > Slack test messages > signals milestone eligibility around the real Slack account link and stays quiet on reconfirmation": 977,
|
||||
"automatic Slack registration > uploads a real 512-pixel PNG from the production renderer": 2228,
|
||||
"automatic Slack registration > uploads preset bytes after credentials are durable, even with private avatar %s": 446,
|
||||
"automatic Slack registration > uses an authenticated message event to complete automatic setup without a URL challenge and notifies once": 889,
|
||||
"automatic Slack registration > vaults the bot token before %s and resumes without code exchange": 1542,
|
||||
"automatic Slack registration > wakes verification from a signed callback with no maintenance polling": 846,
|
||||
"automatically enables native modals for an existing Discord endpoint after %s runtime qualification": 1364,
|
||||
"automatically recovers a pre-ingress GitHub failure only through a genuine callback and normal durable workers": 976,
|
||||
"binds Slack DM slash controls to the latest task ($control, $status, $state)": 5392,
|
||||
"binds Slack tools to the admitted linked request, recovers its provenance and revokes retained access": 4776,
|
||||
"blocks a committed native review response and selected files at transport after %s": 2645,
|
||||
"Board send terminal attachment rejection > commits one content-free scoped receipt for concurrent rejection and allows only a separately keyed corrected send": 971,
|
||||
"Board send terminal attachment rejection > does not classify missing attachments as a safe terminal rejection": 824,
|
||||
"Board send terminal attachment rejection > keeps the exact rejected key terminal after normal owner-comment deletion and service reconstruction": 894,
|
||||
"Board send terminal attachment rejection > returns the exact HTTP rejection with only the bound subset and replays that receipt unchanged": 808,
|
||||
"bounds a held Slack receipt before delivering an already-ready same-endpoint final": 2450,
|
||||
"bounds company fixture prefix collision retries without creating a partial fixture": 86,
|
||||
"bounds failed retry notice scope discovery with retained candidates (%s)": 2526,
|
||||
"bounds inbound notice scope discovery with denied history (%s)": 2739,
|
||||
"bounds interaction continuation recovery reads with completed history": 2767,
|
||||
"bounds terminal confirmation recovery reads with completed history": 3114,
|
||||
"cancels a queued Slack slash start when channel reach is revoked before transport": 315,
|
||||
"cancels and redacts GitHub ingress when rotation wins before replay authentication": 823,
|
||||
"cancels staged GitHub ingress when a minimal recovery tombstone arrives before processing": 660,
|
||||
"cancels staged GitHub ingress when same-secret pause and resume wins at the replay barrier": 662,
|
||||
"cannot publish a runtime whose initialization is overtaken by pause": 308,
|
||||
"canonicalizes prefixed Chat SDK channel ids onto provider inventory resources": 346,
|
||||
"captures initial Slack communication guidance once per task, ignoring forged message configuration": 787,
|
||||
"captures signed Slack forwarding protocol with registered owners (%s)": 1698,
|
||||
"checks both linked sender and agent assignment grants for protected organizational defaults": 644,
|
||||
"checks protected organizational defaults against the assignee locked at creation": 341,
|
||||
"classifies publication $label": 5195,
|
||||
"cleans only the GitHub final run's receipt and never re-adds it on replay": 830,
|
||||
"clears an older multi-line Slack receipt diagnostic on terminal removal with attachment warning %s": 893,
|
||||
"coalesces closed native progress on %s fixture %s (%s) without projecting event content": 5344,
|
||||
"coalesces concurrent automatic pre-ingress GitHub scans into one provider redelivery": 611,
|
||||
"coalesces one Microsoft Teams run into one provider reply": 824,
|
||||
"coalesces one Slack run's lifecycle and final response despite an interleaved task control": 874,
|
||||
"coalesces one Telegram run into one provider message": 1194,
|
||||
"coalesces Telegram status and final output into one run-scoped provider message": 650,
|
||||
"composes a parsed Discord question answer with one same-thread continuation publication": 1185,
|
||||
"configures a customer-owned GitHub App and only auto-enables the first addressed setup repository": 1100,
|
||||
"configures a customer-owned Microsoft Teams bot with the entered credentials": 158,
|
||||
"configures a Discord bot only after identity, intent, server, and channel permissions verify": 1300,
|
||||
"configures Telegram and preserves queued updates with webhook origin %s": 1487,
|
||||
"consumes Telegram forum-topic controls without remapping or waking the topic task": 645,
|
||||
"continues a legacy Teams binding across a regional route change without forking its task": 695,
|
||||
"continues an accepted GitHub task Retry from recovered saved work when its chat replay is refused after the click: %s": 50877,
|
||||
"continues an inherited GitHub task from its recorded result instead of replaying the committed provider failure": 964,
|
||||
"corrects Telegram DM conversation links without rewriting retained provider receipts": 568,
|
||||
"creates a channel-purpose connection and rejects attempts to change its assigned agent": 132,
|
||||
"deduplicates inbound events, keeps one task per thread, and enables newly discovered Slack channels": 887,
|
||||
"deduplicates Slack file-only revisions by consumed metadata without retaining private locators": 442,
|
||||
"defers a rate-limited Telegram receipt reaction for the full provider interval without blocking the inbound turn": 310,
|
||||
"defers continuation dispatch until the outer response transaction commits": 1692,
|
||||
"defers scheduled accepted-response recovery behind %s status effects": 931,
|
||||
"defers scheduled accepted-response recovery until the persisted retry is due": 614,
|
||||
"defers Slack file receipt recovery while the original post owns the same attempt": 827,
|
||||
"delivers and replaces a Slack %s card through the publication service": 5354,
|
||||
"delivers oversized Discord Markdown losslessly as one retryable attachment": 1552,
|
||||
"delivers queued provider notices from commit wakes without the deferred fast path": 307,
|
||||
"denies a pre-provider reviewed retry with %s": 8256,
|
||||
"denies a staged failed chat retry after its original lifecycle epoch changes": 564,
|
||||
"denies a stale automatic GitHub callback after pause and reconnect even with the still-current HMAC secret": 681,
|
||||
"denies guest Slack stops, fences delayed events from newer runs, and cancels queued wakes": 1244,
|
||||
"Discord automatic command registration > does not rediscover a receiver stopped by a company removal refresh": 391,
|
||||
"Discord automatic command registration > drives receiver discovery through real configure, rotation, pause, and resume producers": 1114,
|
||||
"Discord automatic command registration > fences receiver construction across renewal: %s": 6197,
|
||||
"Discord automatic command registration > keeps mention setup usable on registration failure and upgrades automatically when due": 622,
|
||||
"Discord automatic command registration > learns the first configured command and cancels its deadline when the endpoint is removed": 551,
|
||||
"Discord automatic command registration > preserves a healthy Discord Gateway when a due command refresh returns 503": 524,
|
||||
"Discord automatic command registration > rechecks durable ownership after an earlier renewal finishes before listener start": 502,
|
||||
"Discord automatic command registration > reconstructs an uncertain registration by GET without repeating the provider write": 546,
|
||||
"Discord automatic command registration > restarts a receiver after token loss only after the old listener has stopped": 2448,
|
||||
"Discord automatic command registration > retains the Gateway after an external command namespace conflict without overwriting it": 534,
|
||||
"Discord automatic command registration > takes over a persisted foreign receiver lease at its deadline through the worker": 1515,
|
||||
"Discord automatic command registration > updates scheduled registration after real %s": 1448,
|
||||
"Discord native commands with durable service authority > acknowledges privately before a blocked DB wait but denies %s before commit": 1925,
|
||||
"Discord native commands with durable service authority > answers a cold unavailable Discord command privately without leaving the interaction thinking": 867,
|
||||
"Discord native commands with durable service authority > denies a previously accepted command after %s without another action or public receipt": 7514,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > allows a complete freshly admitted coalesced batch": 1510,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > allows an admitted source sent after close command but before close confirmation to present": 1536,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > does not cross a historical close after current %s authority changes": 13914,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > does not drop an older coalesced source while selecting the fresh batch": 1557,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > does not grant presentation across a published control with %s proof": 4611,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > does not let an automatic child copy %s presentation authority": 2049,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > does not let an unseen stale native Discord %s act on reopened work": 2800,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > keeps a newly closed source internal after a still newer source reopens the thread": 1693,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > keeps a source admitted before close confirmation internal after a fresh source reopens": 1421,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > orders same-millisecond Discord sources only within the same worker and process": 1314,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > publishes working and the exact fresh final after pinned close without resurrecting the old source": 1666,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > refuses %s provider chronology at intake when first ingested after close": 2643,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > wakes a post-close milestone when %s authority is restored": 24541,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > wakes a post-close milestone when interrupted inbound acceptance settles its %s receipt": 4707,
|
||||
"Discord native commands with durable service authority > fresh-after-close publication bindings > wakes saved post-close native progress when %s authority returns": 10977,
|
||||
"Discord native commands with durable service authority > qualifies pinned Discord close-owned progress retirement (%s)": 4434,
|
||||
"Discord native commands with durable service authority > records DM new once and never applies its old replay to a replacement conversation": 1117,
|
||||
"Discord native commands with durable service authority > returns private status, keeps guild new as guidance, and closes only after a real public control receipt": 1018,
|
||||
"Discord native commands with durable service authority > rolls back staged public control when the durable command receipt fails": 784,
|
||||
"Discord native commands with durable service authority > serializes simultaneous exact close deliveries and denies a changed command or registered ID": 859,
|
||||
"Discord native commands with durable service authority > upgrades a retained prior-copy registration after service restart before admitting close": 985,
|
||||
"discovers a Slack connect identity without starting work or granting access": 534,
|
||||
"dispatches and publishes a queued Slack Board send with a one-connection database pool": 680,
|
||||
"dispatches saved input through the task Retry HTTP endpoint and survives a repeated click": 1261,
|
||||
"does not accept a contradictory HTTP%s Slack duplicate receipt": 729,
|
||||
"does not acknowledge lifecycle callbacks whose durable write fails": 287,
|
||||
"does not adopt historical pre-ingress GitHub failures when initializing a recovery epoch": 480,
|
||||
"does not adopt input or deliver a pending identity-contended Retry after %s": 5471,
|
||||
"does not bypass %s task guards from the Slack Board composer or its outbox": 2844,
|
||||
"does not bypass %s through an exact chat retry": 1932,
|
||||
"does not cancel an ambiguous Slack task start after credential-lease ownership is lost": 48,
|
||||
"does not commit a Telegram close control when its provider reply is definitively rejected": 860,
|
||||
"does not commit endpoint reach changes after credential-lease ownership is lost": 42,
|
||||
"does not commit stale Slack task-start recovery after credential-lease ownership is lost": 72,
|
||||
"does not complete GitHub webhook repair or overwrite a new credential lease owner after remote success": 553,
|
||||
"does not create SDK eyes reactions for accepted GitHub requests": 591,
|
||||
"does not deadlock reaction replay against publication link settlement": 882,
|
||||
"does not dispatch an accepted inbound wake after destination access is revoked": 415,
|
||||
"does not duplicate an inbound file when delivery recovery resumes after the task mutation": 585,
|
||||
"does not enable Discord modals after a connection-lock wait with %s": 731,
|
||||
"does not hide another endpoint behind a page of busy-bot conversation heads": 2995,
|
||||
"does not hold another endpoint's final publication behind Slack session status transport": 1065,
|
||||
"does not lend a consumed failure lane to a recovered answer with %s": 2541,
|
||||
"does not let a late %s milestone replace authoritative ownership attention or final status": 4033,
|
||||
"does not let an obsolete GitHub webhook secret reopen a repository during rotation": 775,
|
||||
"does not let an obsolete Slack signing secret verify a reconnected configuration": 446,
|
||||
"does not let an obsolete Slack task-start failure quarantine successor credentials or reach": 283,
|
||||
"does not let GitHub lifecycle recovery undo a webhook-secret rotation": 793,
|
||||
"does not let ineligible pre-ingress GitHub candidates exhaust every future recovery scan": 595,
|
||||
"does not let one blocked Slack conversation starve another outbox head": 624,
|
||||
"does not let pause overtake a credential-fenced Discord root thread creation": 284,
|
||||
"does not let pause return while a Slack publication is still in provider transport": 864,
|
||||
"does not let rejected callbacks or malformed proxy hints establish healthy Slack callbacks": 490,
|
||||
"does not let removed-input cleanup overwrite a pre-run FIFO lane already replaced by its answer": 834,
|
||||
"does not let Slack reach disables return ahead of in-flight provider sends": 1164,
|
||||
"does not let stale GitHub setup-secret rotation overwrite successor credentials": 118,
|
||||
"does not let stale Telegram removal clear successor credentials after lease reclamation": 227,
|
||||
"does not let two live agent connections claim the same native bot or spoof its verified identity": 175,
|
||||
"does not lock durably ineligible %s history on repeated issue milestones": 17979,
|
||||
"does not normalize a stale Slack retry-anyway action after credential-lease ownership is lost": 251,
|
||||
"does not recover a committed answer after %s": 6502,
|
||||
"does not reload dormant accepted-response history for an unrelated %s wake": 3138,
|
||||
"does not repeat a failed chat run with %s": 1066,
|
||||
"does not replace a failure while another exact-run authored publication is %s": 1460,
|
||||
"does not replay stale native progress after a paused endpoint restarts behind an exact final": 1596,
|
||||
"does not request automatic pre-ingress GitHub recovery while the latest provider attempt is %s": 1899,
|
||||
"does not resurrect publication ownership after waiting beyond expiry on a real lease row lock": 464,
|
||||
"does not retain signed GitHub events outside the configured App event set": 510,
|
||||
"does not retry a chat comment edited on the Board during the failed execution": 362,
|
||||
"does not retry a provider-accepted publication after its exact attempt lease is lost": 459,
|
||||
"does not retry automatic pre-ingress GitHub redelivery without a distinct failed provider attempt": 569,
|
||||
"does not retry unrelated company fixture database errors": 10,
|
||||
"does not rotate a GitHub webhook secret when stored credentials cannot be resolved": 470,
|
||||
"does not treat an accepted-response endpoint scheduling hint as current authority": 512,
|
||||
"does not wake twice when an operator replays an already admitted failed delivery": 478,
|
||||
"drains accepted GitHub webhooks from commit wakes and durable retry deadlines": 2018,
|
||||
"drains bounded Slack chat-origin milestones without admitting an internal issue run (%s)": 2465,
|
||||
"drains failed chat retries from durable work and becomes idle: %s": 9768,
|
||||
"durably admits a Telegram edit that races deferred original-message processing": 1181,
|
||||
"durably audits a rejected Discord Gateway action before surfacing transport rejection": 225,
|
||||
"durably deduplicates Telegram guidance webhook retries and filters disabled DMs": 384,
|
||||
"durably recovers rate-limited Telegram menu registration and removal cleanup": 674,
|
||||
"durably replays a reaction that arrives before its outbound link without waking the task": 991,
|
||||
"durably retries a signed Slack session stop and cancels its exact linked run once": 1025,
|
||||
"durably stages GitHub webhooks while credential mutation owns the processing lease": 880,
|
||||
"durably stages GitHub webhooks within budget before cold runtime initialization": 629,
|
||||
"edits a Slack working reply into one terminal failure reply": 581,
|
||||
"enforces Telegram group reach and privacy-mode addressing before retaining message content": 674,
|
||||
"enforces the native progress cadence boundary and one publication per phase": 692,
|
||||
"exhausts the existing delivery Retry budget after identity authorization without background renewal": 8538,
|
||||
"expires Teams consent from the committed Board transfer without periodic maintenance": 1988,
|
||||
"expires the saved identity wait without renewal when the lock is released before expiry processing: %s": 1807,
|
||||
"fails closed for unbound rich callbacks and unknown slash commands": 279,
|
||||
"fails closed on automatic pre-ingress GitHub recovery with %s": 980,
|
||||
"fails closed when GitHub App permissions drift before installation recovery": 1295,
|
||||
"fails closed when the GitHub installation has not approved the App permissions": 859,
|
||||
"fences a reclaimed Slack task admission from the obsolete worker's settlement": 411,
|
||||
"fences message and reaction callbacks that resume after Discord Gateway takeover": 1258,
|
||||
"filters a claimed message lifecycle delivery after pause and reconnect supersede its runtime": 915,
|
||||
"filters a durable orphan Telegram edit after the bounded reorder window": 555,
|
||||
"filters a durably admitted GitHub callback when rotation and reconnect supersede its runtime before deferred processing": 947,
|
||||
"filters a late duplicate instead of bypassing the bounded reaction-link lifetime": 1024,
|
||||
"filters a staged reaction when destination reach is revoked before replay": 929,
|
||||
"filters a standalone unaddressed GitHub comment after bounded orphan retention": 943,
|
||||
"filters an unlinked guest when the endpoint sponsor is suspended and exposes the reason and duplicate count": 313,
|
||||
"filters unlinked people when configured, then honors a confirmed active-member identity link": 827,
|
||||
"first-seen messages behind completed chat controls > accepts delayed %s backlog without an explicit control": 1734,
|
||||
"first-seen messages behind completed chat controls > continues to accept authenticated Slack slash close/new on a genuinely new generation": 1841,
|
||||
"first-seen messages behind completed chat controls > does not infer a stop from bare completed state with %s": 3450,
|
||||
"first-seen messages behind completed chat controls > does not let an unseen old Telegram /%s close a newer generation": 2071,
|
||||
"first-seen messages behind completed chat controls > does not lower a newer provider boundary when an older close is confirmed later": 1420,
|
||||
"first-seen messages behind completed chat controls > does not promote metadata-only %s timestamps after a completed control": 1986,
|
||||
"first-seen messages behind completed chat controls > does not recreate or filter a previously committed inbound link when its delivery retries after close": 1000,
|
||||
"first-seen messages behind completed chat controls > honors operator-confirmed close without inventing an outbound message receipt": 930,
|
||||
"first-seen messages behind completed chat controls > keeps controls scoped to their endpoint, immutable thread, and company": 2827,
|
||||
"first-seen messages behind completed chat controls > preserves a Teams close across regional routes (close after route change: %s)": 1943,
|
||||
"first-seen messages behind completed chat controls > rechecks a close committed while old input waits before the endpoint transaction": 865,
|
||||
"first-seen messages behind completed chat controls > refuses unseen %s input before %s (same second: %s) without creating work": 8159,
|
||||
"first-seen messages behind completed chat controls > retains a committed control boundary through pause, redacted duplicate, and resume": 1909,
|
||||
"first-seen messages behind completed chat controls > retains a completed trailing-newline control as a first-seen source boundary": 888,
|
||||
"GitHub agent review workflow > acknowledges a GitHub request after a durable scheduler retry succeeds": 1181,
|
||||
"GitHub agent review workflow > adds the progress tool to an existing enabled bot without restoring disabled tools": 1018,
|
||||
"GitHub agent review workflow > admits a GitHub request explicitly addressed to both connected bots": 1583,
|
||||
"GitHub agent review workflow > admits a newly mentioned %s edit while preserving its lifecycle record": 3264,
|
||||
"GitHub agent review workflow > admits opt-in signed issue events without PR checks, deduplicates them and rechecks revoked policy": 1684,
|
||||
"GitHub agent review workflow > admits signed %s mentions with automation off, deduplicates and retains manual task authority": 33894,
|
||||
"GitHub agent review workflow > admits signed PR events into an ordinary assigned task and deduplicates delivery": 4468,
|
||||
"GitHub agent review workflow > adopts an existing GitHub PR placeholder and retires earlier duplicate placeholders": 2233,
|
||||
"GitHub agent review workflow > authorizes description edits as the editor, never inheriting the original author's access": 2508,
|
||||
"GitHub agent review workflow > automatically retries a transient GitHub review publication (HTTP %s)": 9403,
|
||||
"GitHub agent review workflow > bounds GitHub response authorization reads with large saved chat content": 1359,
|
||||
"GitHub agent review workflow > checks live repository permission and reports a revoked installation selection": 559,
|
||||
"GitHub agent review workflow > classifies recovered GitHub receipts without republishing or overwriting historical success": 2391,
|
||||
"GitHub agent review workflow > closes a GitHub placeholder after a formal review (legacy receipt: %s)": 2546,
|
||||
"GitHub agent review workflow > closes an inline-only GitHub review when an earlier working comment was deleted: %s": 2994,
|
||||
"GitHub agent review workflow > coordinates GitHub working feedback with a %s fast response": 3605,
|
||||
"GitHub agent review workflow > defers GitHub review check enqueue until outer commit and discards rollback": 1878,
|
||||
"GitHub agent review workflow > does not admit an unlinked issue author through the linked webhook sender": 648,
|
||||
"GitHub agent review workflow > does not edit a %s replacement of the bound GitHub comment": 2524,
|
||||
"GitHub agent review workflow > does not handle a blocked new GitHub request with an older run's final reply": 1728,
|
||||
"GitHub agent review workflow > does not invent a GitHub completion response when the agent sends no tool reply": 1418,
|
||||
"GitHub agent review workflow > does not let a progress edit overwrite a final GitHub answer": 1475,
|
||||
"GitHub agent review workflow > does not let an unsettled GitHub fallback block a later board message": 1139,
|
||||
"GitHub agent review workflow > does not post a GitHub working comment when the scheduler declines the request": 671,
|
||||
"GitHub agent review workflow > does not resubmit a permanent GitHub publication rejection (HTTP %s)": 3371,
|
||||
"GitHub agent review workflow > does not route GitHub mentions using %s identities": 4374,
|
||||
"GitHub agent review workflow > drains more than one page of stale GitHub review projections without touching manual history": 2133,
|
||||
"GitHub agent review workflow > finalizes every working comment in a coalesced GitHub run": 2064,
|
||||
"GitHub agent review workflow > finds a GitHub review check after a full page of revoked work and ignores settled history": 6083,
|
||||
"GitHub agent review workflow > finds owed terminal GitHub work beyond a full page of unrelated request history": 1957,
|
||||
"GitHub agent review workflow > finds terminal GitHub responses by native issue binding before legacy run context": 2137,
|
||||
"GitHub agent review workflow > finishes the same GitHub working comment after an authorized same-App reconnect (legacy: %s)": 2594,
|
||||
"GitHub agent review workflow > GitHub failure fallback respects exact tool delivery state: %s": 14536,
|
||||
"GitHub agent review workflow > holds stale GitHub review projection receipt repair until the outer commit and discards rollback": 2793,
|
||||
"GitHub agent review workflow > ignores a review request addressed to an external GitHub bot": 1023,
|
||||
"GitHub agent review workflow > imports initial GitHub access for a %s App without a second picker and leaves later additions disabled": 1597,
|
||||
"GitHub agent review workflow > keeps a newer GitHub request open when its acknowledgement follows older progress": 2279,
|
||||
"GitHub agent review workflow > keeps a report-only GitHub check neutral after a %s run without an assessment": 2930,
|
||||
"GitHub agent review workflow > keeps an acknowledged paused GitHub request quiet until its new run finishes": 2862,
|
||||
"GitHub agent review workflow > keeps GitHub publication I/O outside transactions and fences an expired owner's receipt": 422,
|
||||
"GitHub agent review workflow > keeps GitHub response maintenance truthful for a new request: $scenario ($reason)": 54805,
|
||||
"GitHub agent review workflow > keeps GitHub response ownership checks independent of large run payloads": 1422,
|
||||
"GitHub agent review workflow > keeps GitHub review check history quiet and wakes from the real enqueue producer": 2895,
|
||||
"GitHub agent review workflow > keeps GitHub review tools callable through a later native task gateway (board follow-up: $boardFollowUp, legacy origin: $legacyOrigin)": 9341,
|
||||
"GitHub agent review workflow > keeps native presentation idle until its committed GitHub outcome: %s": 5965,
|
||||
"GitHub agent review workflow > keeps partial review delivery authoritative when the head changes %s": 2589,
|
||||
"GitHub agent review workflow > keeps per-request fallback outside the main GitHub PR review wait": 1361,
|
||||
"GitHub agent review workflow > keeps terminal GitHub responses waiting while a natively bound newer run is active": 871,
|
||||
"GitHub agent review workflow > legacy name-only registration preserves an explicit empty repository selection": 98,
|
||||
"GitHub agent review workflow > links a gated GitHub check to the current vanity review page before a task exists": 731,
|
||||
"GitHub agent review workflow > lists each GitHub bot tool once when an inline thread shares the task": 1155,
|
||||
"GitHub agent review workflow > loads an old GitHub review directly without crossing endpoint or company boundaries": 1390,
|
||||
"GitHub agent review workflow > loads only the current configured instruction skills, never provider links or another company": 2562,
|
||||
"GitHub agent review workflow > manual App credential retry imports repositories after an interrupted vault write": 1201,
|
||||
"GitHub agent review workflow > manual App setup preserves explicitly disabled repositories and tools": 789,
|
||||
"GitHub agent review workflow > manual App setup resumes tool initialization after an interrupted repository import": 929,
|
||||
"GitHub agent review workflow > offers Retry for a saved automatic PR event after a successful older review": 2910,
|
||||
"GitHub agent review workflow > parks a GitHub acknowledgement until native presentation commits": 1352,
|
||||
"GitHub agent review workflow > parks terminal GitHub responses behind the publication owner (uncertain: %s)": 6911,
|
||||
"GitHub agent review workflow > preserves %s policy context while omitting body content": 1779,
|
||||
"GitHub agent review workflow > preserves a final GitHub assessment while settling an incomplete request receipt": 1392,
|
||||
"GitHub agent review workflow > preserves an earlier final GitHub answer when a later continuation fails": 1132,
|
||||
"GitHub agent review workflow > preserves GitHub connection intent after publication HTTP %s": 2572,
|
||||
"GitHub agent review workflow > preserves malformed GitHub authority input %j": 5336,
|
||||
"GitHub agent review workflow > preserves newer GitHub PR progress when an older run fails": 1874,
|
||||
"GitHub agent review workflow > preserves the %s GitHub receipt lifecycle during acknowledgement recovery": 2360,
|
||||
"GitHub agent review workflow > preserves the published GitHub assessment during passive-response maintenance: %s": 4814,
|
||||
"GitHub agent review workflow > publishes one score heading when a newly configured bot submits a formatted summary": 1880,
|
||||
"GitHub agent review workflow > reactivates terminal GitHub response recovery from committed native identity": 3258,
|
||||
"GitHub agent review workflow > reads pending GitHub receipts without dispatch and refuses a missing publication": 2999,
|
||||
"GitHub agent review workflow > rechecks a GitHub fallback after waiting for the credential lane": 1231,
|
||||
"GitHub agent review workflow > rechecks another GitHub bot's mention when recovering a durable subscribed delivery": 1565,
|
||||
"GitHub agent review workflow > rechecks assessment ownership after stale GitHub review projection selection waits for a row lock (shutdown: %s)": 3635,
|
||||
"GitHub agent review workflow > reconciles a GitHub working comment when execution stops without a publication": 1726,
|
||||
"GitHub agent review workflow > reconciles stopped GitHub work from native identity without legacy wake fields": 1287,
|
||||
"GitHub agent review workflow > reconsiders the same GitHub reply after stopped native workspace cleanup (%s)": 40213,
|
||||
"GitHub agent review workflow > records automatic policy skips without retaining provider content, and refuses mentions in disabled repositories": 866,
|
||||
"GitHub agent review workflow > recovers %s after a timeout before publication staging": 3134,
|
||||
"GitHub agent review workflow > recovers a delayed GitHub acknowledgement after a successful run without leaving it working": 1216,
|
||||
"GitHub agent review workflow > recovers a durable GitHub publication receipt ($label)": 15195,
|
||||
"GitHub agent review workflow > recovers a missed GitHub head webhook only while a review check remains unresolved": 1087,
|
||||
"GitHub agent review workflow > recovers an acknowledgement whose local receipt was lost without posting another comment": 1190,
|
||||
"GitHub agent review workflow > recovers an approved queued GitHub review without another approval (revoked: %s)": 5397,
|
||||
"GitHub agent review workflow > recovers long %s requests without accepting changes hidden by the audit preview": 7503,
|
||||
"GitHub agent review workflow > recovers lost inline and check responses without duplicate GitHub publications": 9067,
|
||||
"GitHub agent review workflow > recovers stale GitHub review projection through %s": 4436,
|
||||
"GitHub agent review workflow > rejects arbitrary comment IDs and edits after the initiating person is revoked": 1039,
|
||||
"GitHub agent review workflow > rejects changed requests and revoked access before repeating an unstaged review": 7706,
|
||||
"GitHub agent review workflow > releases a GitHub acknowledgement after fast response %s without maintenance polling": 9886,
|
||||
"GitHub agent review workflow > replies in review finding threads behind an active run > answers each finding thread in its own run instead of one unattestable batch": 6263,
|
||||
"GitHub agent review workflow > replies in review finding threads behind an active run > still batches several replies in one finding thread into one attested run": 4161,
|
||||
"GitHub agent review workflow > replies in review finding threads behind an active run > still denies each separated finding-thread run after the linked identity is revoked": 2688,
|
||||
"GitHub agent review workflow > requires every essential GitHub tool and rejects Contents write during verification": 1148,
|
||||
"GitHub agent review workflow > restores terminal GitHub responses and leaves finalized history quiet": 1913,
|
||||
"GitHub agent review workflow > resumes a delayed GitHub publication after coordinator restart without another inbound event": 3487,
|
||||
"GitHub agent review workflow > resumes a paused GitHub request and reconciles stopped execution (prior run: %s, progress: %s)": 8763,
|
||||
"GitHub agent review workflow > retains a provider-verified GitHub mention when its plain text uses an alias": 1072,
|
||||
"GitHub agent review workflow > retains attachment provenance in a signed GitHub mention %s": 2722,
|
||||
"GitHub agent review workflow > retains failed stale GitHub review projection repair across drain without passive SQL": 1788,
|
||||
"GitHub agent review workflow > retains GitHub review check work across %s": 7576,
|
||||
"GitHub agent review workflow > retains subscribed GitHub follow-ups with incidental human or unknown mentions: %s": 7392,
|
||||
"GitHub agent review workflow > retains terminal GitHub response work across %s": 6174,
|
||||
"GitHub agent review workflow > retains the saved GitHub review check retry through restart without an early provider call": 1053,
|
||||
"GitHub agent review workflow > retires a legacy coalesced GitHub placeholder before replacing its receipt": 2942,
|
||||
"GitHub agent review workflow > retires a legacy GitHub placeholder when its acknowledgement is superseded": 2048,
|
||||
"GitHub agent review workflow > retries a failed GitHub acknowledgement at its saved deadline": 1842,
|
||||
"GitHub agent review workflow > retries an ambiguous GitHub edit without creating another comment": 7774,
|
||||
"GitHub agent review workflow > returns snapshot finding categories and publishes only accepted categories": 2307,
|
||||
"GitHub agent review workflow > reuses one main GitHub PR comment across review requests: %s": 5655,
|
||||
"GitHub agent review workflow > scopes guest tasks and rechecks their sponsor before every bot tool": 875,
|
||||
"GitHub agent review workflow > settles a deleted Board-origin request without starving unrelated terminal GitHub replies": 3066,
|
||||
"GitHub agent review workflow > settles a deleted coalesced working comment without blocking the current GitHub response": 1410,
|
||||
"GitHub agent review workflow > stack exchanges a callback code only once when vault storage is interrupted": 54,
|
||||
"GitHub agent review workflow > stack rejects forged raw gateway deliveries without changing verification or waking an agent": 1018,
|
||||
"GitHub agent review workflow > stages one assessment when the timed-out metadata read completes alongside its retry": 2522,
|
||||
"GitHub agent review workflow > still fences a GitHub publication when reconnect races the provider read": 1203,
|
||||
"GitHub agent review workflow > stops a queued GitHub publication after %s": 5732,
|
||||
"GitHub agent review workflow > stops GitHub acknowledgement sends at %s between provider requests": 5969,
|
||||
"GitHub agent review workflow > stops GitHub review check HTTP after shutdown admission and recovers its durable action": 7325,
|
||||
"GitHub agent review workflow > supersedes a late review projection after its check already rejected the old head": 1769,
|
||||
"GitHub agent review workflow > suppresses subscribed GitHub comments for another connected bot in %s": 4035,
|
||||
"GitHub agent review workflow > updates one GitHub acknowledgement through progress and final answer in %s": 5490,
|
||||
"GitHub agent review workflow > uses task-bound bot tools and deterministic checks, then denies revoked people": 3438,
|
||||
"GitHub agent review workflow > uses the manifest state parameter and rejects expired, reused, and wrong-origin registrations": 53,
|
||||
"GitHub agent review workflow > verifies %i enabled repositories in bounded token batches": 1136,
|
||||
"GitHub agent review workflow > verifies a repaired GitHub webhook in %s only from a %s installation delivery with %s repository": 7495,
|
||||
"GitHub agent review workflow > waits for and wakes from the actual GitHub review check publication lease": 1853,
|
||||
"GitHub agent review workflow > waits for the existing GitHub publication lease before a terminal response": 1951,
|
||||
"GitHub agent review workflow > wakes a GitHub acknowledgement only after the outer fast-response transaction commits": 2398,
|
||||
"GitHub agent review workflow > wakes a GitHub review check from committed run identity before its future head deadline": 2749,
|
||||
"GitHub agent review workflow > wakes a paused GitHub response from %s": 10137,
|
||||
"GitHub agent review workflow > wakes GitHub acknowledgements from a real publication lease %s": 5736,
|
||||
"GitHub agent review workflow > wakes GitHub milestone identity restoration only after outer commit and stays quiet on reconfirmation": 424,
|
||||
"GitHub agent review workflow > wakes milestone authority only when GitHub guest settings change, after outer commit": 657,
|
||||
"GitHub agent review workflow > wakes stale GitHub review projection repair from real admission and leaves settled history quiet": 1816,
|
||||
"GitHub agent review workflow > wakes stale GitHub review projection repair when cancellation commits after the review is admitted": 2517,
|
||||
"GitHub agent review workflow > wakes terminal GitHub responses after the outer issue transaction commits": 2037,
|
||||
"GitHub agent review workflow > wizard completes a verified connection automatically without claiming runtime execution": 762,
|
||||
"GitHub agent review workflow > wizard configures Cloud webhook delivery without a public instance URL": 528,
|
||||
"GitHub agent review workflow > wizard keeps localhost draft choices while waiting for Cloud enrollment": 53,
|
||||
"GitHub agent review workflow > wizard persists a %s owner and resumes the same bound registration after a lost Cloud response and enrolled origin change": 172,
|
||||
"GitHub agent review workflow > wizard preserves an explicit empty repository selection at revision zero": 679,
|
||||
"GitHub agent review workflow > wizard preserves explicit review behavior and repository restrictions on existing drafts": 671,
|
||||
"GitHub agent review workflow > wizard recovers a renamed enrolled origin only after checking bound state and Cloud authority": 52,
|
||||
"GitHub agent review workflow > wizard repairs Cloud after manual credential recovery and secret rotation without changing App identity": 824,
|
||||
"GitHub agent review workflow > wizard restarts only an expired unconsumed registration: %s": 490,
|
||||
"GitHub agent review workflow > wizard resumes a %s installation callback after its signed ping": 2896,
|
||||
"GitHub agent review workflow > wizard retains partial manifest recovery until explicit existing-App recovery succeeds": 941,
|
||||
"GitHub agent review workflow > wizard serializes identity starts and callbacks without overwriting PKCE or handoff state": 603,
|
||||
"GitHub repository pagination and bulk access > pages and searches 1000 repositories while toggle all changes unloaded rows atomically": 425,
|
||||
"GitHub terminal receipt composes pinned HTTP and durable source authority (%s)": 15635,
|
||||
"GitHub webhook recovery scheduling > joins held GitHub history at shutdown without starting another HTTP request": 565,
|
||||
"GitHub webhook recovery scheduling > leaves %s recovery quiet": 971,
|
||||
"GitHub webhook recovery scheduling > preempts a future GitHub recovery window when only credential references change": 1465,
|
||||
"GitHub webhook recovery scheduling > preserves a GitHub redelivery stopped before transport (%s, prior receipt: %s, superseded: %s)": 3904,
|
||||
"GitHub webhook recovery scheduling > reactivates recovery from real resume and cancels its deadline on pause": 3506,
|
||||
"GitHub webhook recovery scheduling > reconstructs due GitHub recovery through %s": 1629,
|
||||
"GitHub webhook recovery scheduling > reconstructs more than one page of GitHub recovery deadlines and dispatches fair batches of five": 2602,
|
||||
"GitHub webhook recovery scheduling > recovers a failed GitHub deadline reconstruction after drain without passive SQL": 573,
|
||||
"GitHub webhook recovery scheduling > rejects a %s assigned agent before GitHub recovery can become orphaned": 4431,
|
||||
"GitHub webhook recovery scheduling > retains GitHub recovery configuration intent through nested commit and rollback": 2434,
|
||||
"GitHub webhook recovery scheduling > waits for the saved recovery deadline and requests a genuine GitHub redelivery": 671,
|
||||
"globally fences GitHub App identity across owner transfers without leaking another company": 477,
|
||||
"globally fences one Microsoft Bot application across tenants": 183,
|
||||
"hands an explicitly bound same-run image to Discord's final native file response": 1394,
|
||||
"holds a delayed first Teams setup reply until its older root mention enables the channel": 557,
|
||||
"holds a delayed Slack thread reply until its older root mention arrives": 472,
|
||||
"holds a GitHub follow-up that arrives after the reorder window until its older root mention arrives": 869,
|
||||
"holds ambiguous provider sends for an audited duplicate-risk resolution without reordering": 1465,
|
||||
"holds an ordinary $provider Retry behind its exact remote recovery: $gate": 38604,
|
||||
"honors a prepared Retry identity wait after %s": 9729,
|
||||
"honors Slack slash-start rate-limit backoff without hammering transport": 317,
|
||||
"honors Telegram flood-control timing and publishes one final message after retry": 862,
|
||||
"ignores a signed GitHub installation before App credentials exist": 170,
|
||||
"ignores repeated unscoped routine %s hints without rebuilding dormant accepted-response history": 1027,
|
||||
"ignores signed Slack callbacks from a different workspace before SDK dispatch": 280,
|
||||
"immediately initializes a new GitHub recovery generation despite the previous epoch's 24-hour backoff": 568,
|
||||
"imports every Discord upload when text files include MIME parameters": 424,
|
||||
"ingests a pinned Telegram video-note without optional MIME metadata (%s)": 2277,
|
||||
"ingests bounded Telegram photo, audio, video, and document attachments without stranding text": 305,
|
||||
"ingests only admitted public GitHub attachments after restart (HTTP $status, $count files, expired batch $expireBatch)": 3735,
|
||||
"ingests Teams files only from personal chats and keeps non-DM references link-only": 476,
|
||||
"invalidates only an exact provider-verified GitHub source edit independently of editor admission: %s": 4217,
|
||||
"isolates malformed publication preparation so later messages still deliver": 871,
|
||||
"isolates pre-run FIFO reconciliation scopes and exhausts published history": 686,
|
||||
"joins in-flight recovery before rethrowing an ordinary drain failure (%s)": 2944,
|
||||
"keeps a %s retry's queue, working, and selected final on its own provider identity": 1778,
|
||||
"keeps a board message queued separately when saved GitHub Retry resumes after cleanup": 975,
|
||||
"keeps a chat-origin run presentation internal when its external question owns the reply": 1002,
|
||||
"keeps a claimed $provider identity immutable when first setup resumes from attention": 96,
|
||||
"keeps a first-channel Teams orphan payload-free during grace and filters it without a live retry": 209,
|
||||
"keeps a refused GitHub task Retry continued when reconciliation read it before the refusal settled": 729,
|
||||
"keeps a saved HTTP Retry alive through a healthy nine-second identity transaction": 11016,
|
||||
"keeps a successful GitHub link-question continuation final internal": 1344,
|
||||
"keeps a verified Slack $kind authoritative for $contentType read and reuse after reach or identity returns": 9182,
|
||||
"keeps accepted-response work behind the outer %s boundary": 3907,
|
||||
"keeps an aged live publication owned across standby sweeps but quarantines only exact-lease orphans": 851,
|
||||
"keeps an internal Telegram run summary private and completes its progress message": 1048,
|
||||
"keeps an omitted-file warning after an early Slack receipt failure and %s retry": 851,
|
||||
"keeps bare Slack mentions subject to %s access restrictions": 496,
|
||||
"keeps case-variant Teams Entra object ids on one external principal": 390,
|
||||
"keeps current GitHub task Retry authority gated: %s": 11935,
|
||||
"keeps disabled-channel Slack %s scoped to its exact source without admitting content": 2078,
|
||||
"keeps external-continuation lifecycle comments internal and publishes only the selected final": 620,
|
||||
"keeps GitHub issues, PR conversations, and inline review threads on distinct tasks": 1157,
|
||||
"keeps intermediate %s agent comments internal and publishes only the selected final": 1405,
|
||||
"keeps lifecycle and native progress projection independent with %s": 1495,
|
||||
"keeps medium Telegram $label in one native publication": 1482,
|
||||
"keeps missing attachment storage retryable but fails invalid metadata before provider transport": 904,
|
||||
"keeps native progress behind question, final, and current reach authority on %s fixture %s (%s)": 3230,
|
||||
"keeps one durable pre-run FIFO notice across coalescing and restart, then reuses only its successor lane": 1353,
|
||||
"keeps one Paperclip account mapping for the same provider principal across endpoints": 521,
|
||||
"keeps processing audit truth and rejects stale Slack runtime callbacks across resume": 536,
|
||||
"keeps provider startup diagnostics private on %s fixture %s (%s)": 5084,
|
||||
"keeps recovered GitHub final prose and attachment navigation internal (omission: %s)": 1464,
|
||||
"keeps root-chat bookkeeping comments internal and publishes only the selected final": 505,
|
||||
"keeps saved-input Retry behind current admission gates: %s": 4526,
|
||||
"keeps separate Slack DM thread roots on separate tasks": 531,
|
||||
"keeps Slack tools callable after a native task continuation without wake JSON": 763,
|
||||
"keeps successive Telegram DM messages on one active Paperclip task": 812,
|
||||
"keeps task-bound /close active after an ambiguous publication commit until explicit confirmation": 1214,
|
||||
"keeps Teams group chat closed by default without consuming first-channel setup enablement": 340,
|
||||
"keeps Telegram start and unknown commands as terse guidance without creating work": 269,
|
||||
"keeps the admitted chat Retry queued until its source lease cleanup settles: %s": 29364,
|
||||
"keeps the pre-run FIFO sweep running when a valid owner's action has comment ids that are not uuids": 901,
|
||||
"leaves a contended accepted response retryable without reclassifying or partially publishing it (%s)": 1232,
|
||||
"leaves accepted-response presentation to its worker during periodic %s replay": 1208,
|
||||
"leaves Cloud-managed GitHub webhook recovery to the relay": 581,
|
||||
"leaves fresh-commit accepted-response presentation to the worker when periodic replay skips it": 666,
|
||||
"leaves GitHub reconnect unqualified after a %s webhook repair failure": 2321,
|
||||
"leaves saved GitHub input pending when a persisted Retry wait loses board authority": 1145,
|
||||
"links outbound Teams files instead of attempting an unsupported native upload": 1021,
|
||||
"Lossless long Board publications > keeps an ambiguous %s text part blocked across restart and retries only that part and its tail": 5059,
|
||||
"Lossless long Board publications > preserves %s code, Unicode and sanitization expansion through actual durable parts": 6688,
|
||||
"Lossless long Board publications > preserves all 100000 characters for %s %s-comment publication": 39336,
|
||||
"Lossless long Board publications > preserves an accepted native Slack response whose rendered mentions exceed the Markdown cap": 1024,
|
||||
"Lossless long Board publications > refuses malformed persisted wrappers before provider I/O %#": 1678,
|
||||
"Lossless long Board publications > retains a complete 100000-character %s response in the existing Markdown document transport": 1743,
|
||||
"maps a timestamp-bearing Slack DM reaction to the completed generation that owns its linked message": 960,
|
||||
"maps Telegram image, audio, and video output onto native media lanes": 1073,
|
||||
"marks provider setup as needing attention when Telegram webhook registration fails": 136,
|
||||
"materializes direct external-chat finals once and accepts the next turn without agent bookkeeping writes": 3749,
|
||||
"migrates Telegram basic-group reach, topic tasks and $defaultsMode defaults (existing destination: $existingDestination)": 5914,
|
||||
"mirrors Paperclip messages with user attribution and returns only the selected Slack reply": 1127,
|
||||
"never promises a queued follow-up before durable admission succeeds": 448,
|
||||
"never re-arms a failed local GitHub receipt when an automatic redelivery races the original callback": 650,
|
||||
"never replaces or resurrects an already selected answer (deleted=%s)": 1038,
|
||||
"never stores a GitHub webhook body before its signature is authenticated": 428,
|
||||
"opens and corrects an actual Discord modal through current service authority exactly once (%s concurrent answers)": 2722,
|
||||
"orders a Telegram authentication failure before concurrent reconnect": 916,
|
||||
"orders a Telegram destination failure before concurrent membership recovery": 809,
|
||||
"orders a Telegram new command before the next message from the same provider second": 1169,
|
||||
"orders rapid Discord replies by provider time before waking one task": 744,
|
||||
"orders rapid Teams follow-ups by provider time when callbacks arrive reversed": 691,
|
||||
"orders reversed GitHub edit and delete callbacks behind their durable root (root processed before replay: %s)": 6213,
|
||||
"orders Slack private-channel archive, unarchive, and rename lifecycle": 431,
|
||||
"paginates mixed activity beyond 100 rows without losing timestamp ties": 110,
|
||||
"parks pending confirmation history, passes malformed prefixes, and wakes on the outer interaction commit": 9413,
|
||||
"parks provider-confirmed Slack admissions on paused endpoints without starving active work": 665,
|
||||
"passes historical continuation prefixes and retains a response committed during dispatch": 4350,
|
||||
"persists Slack manifest draft details and the registered command, then locks them once connected": 196,
|
||||
"persists Slack membership, uninstall, and same-bot reinstall lifecycle before acknowledging": 807,
|
||||
"persists the first identity wait deadline for an older pending action": 964,
|
||||
"presents a fresh Teams personal generation after close and a regional route change": 888,
|
||||
"preserves a published %s file prefix across restart and retries only an explicitly resolved ambiguous upload": 2533,
|
||||
"preserves a recorded %s presentation instead of recovering an unsolicited answer": 1058,
|
||||
"preserves a Teams personal %s attachment omission after losing its live download closure": 787,
|
||||
"preserves an early Slack receipt failure after %s acceptance": 982,
|
||||
"preserves an explicitly disabled Slack channel after invitation and mentions (%s)": 824,
|
||||
"preserves an interrupted run's selected final when its publication is %s": 1325,
|
||||
"preserves automatic final replies for legacy GitHub connections without saved review configuration": 875,
|
||||
"preserves delivery quarantine when a slow Slack post finishes after stale recovery": 840,
|
||||
"preserves exact Discord source authority through registered Gateway %s updates": 2169,
|
||||
"preserves GitHub webhook verification through setup completion and reconnect": 1071,
|
||||
"preserves one current GitHub Retry across %s": 28187,
|
||||
"preserves ordinary legacy task Retry success without replaying its canonical request": 1959,
|
||||
"preserves running Slack tool authority after Require at-mention is enabled": 469,
|
||||
"preserves selected answer and failure lanes after deferred admission (%s)": 1901,
|
||||
"preserves selected answer and failure lanes on $provider after $status ($order)": 16188,
|
||||
"preserves Slack thread history across completion ($isDM, $threaded, $status, $state)": 2965,
|
||||
"preserves stored Telegram credentials and lifecycle when reconnect uses an unsupported webhook port": 210,
|
||||
"preserves the complete failed coalesced chat batch and rejects a dropped sibling": 599,
|
||||
"preserves the GitHub resource, task, and conversation across repository rename and transfer": 939,
|
||||
"preserves the public inventory error response for %s": 1199,
|
||||
"preserves the raw webhook request and returns the provider adapter response": 323,
|
||||
"preserves unavailable Teams $surface file references through $mode admission": 4337,
|
||||
"projects prepared but unsettled task Retry using its actual stale-claim deadline": 1361,
|
||||
"provider timestamp provenance after closed conversations > retains only raw Telegram date through the pinned parser or runtime slash path (%s)": 3271,
|
||||
"provider timestamp provenance after closed conversations > retains only the actual Teams activity clock, not SDK display fallback (%s)": 4785,
|
||||
"provides assigned Slack tools to routine tasks and can DM only their linked responsible user": 2652,
|
||||
"publishes a closed-choice question, settles its Slack card, and delivers its exact continuation response": 1766,
|
||||
"publishes a serialized Telegram run response after the preceding run completes the conversation": 1068,
|
||||
"publishes a Telegram failure for an exact confirmation continuation that stops before commenting": 1104,
|
||||
"publishes an agent's explicitly selected same-run Slack attachment after its response": 1038,
|
||||
"publishes an explicitly selected board attachment through Telegram's native file lane": 849,
|
||||
"publishes complex question sets as non-executable Paperclip fallbacks": 699,
|
||||
"publishes equal-time Slack outbox rows once in stable order across concurrent drains": 548,
|
||||
"publishes explicit Board GitHub attachment fallbacks without provider file bytes": 1922,
|
||||
"publishes fast responses to the accepted %s thread without completing its work": 1640,
|
||||
"publishes GitHub questions as link-only cards with no executable callback": 558,
|
||||
"publishes long structured Telegram Markdown as one durable lossless document": 1277,
|
||||
"publishes native progress with a typed issue binding present: %s": 1166,
|
||||
"publishes one truthful notice when a selected PNG receives a definite provider rejection": 1054,
|
||||
"publishes only a resolver-selected Telegram reply for $authorizationReason": 2598,
|
||||
"publishes only explicitly selected board attachments and stays idempotent across retries": 1293,
|
||||
"publishes only the safe projection once and locks reassignment of a bound task": 942,
|
||||
"publishes the old final before /new and suppresses later output from the completed generation": 1112,
|
||||
"quarantines a Telegram control when the reply lands but its state commit crashes": 685,
|
||||
"quarantines an active GitHub endpoint when its dedicated App drifts to multiple installations": 851,
|
||||
"quarantines an ambiguous automatic pre-ingress GitHub redelivery without blindly repeating the POST": 568,
|
||||
"quarantines Telegram maintenance success when credential-lease ownership is reclaimed": 605,
|
||||
"queues Telegram guidance on the durable publication FIFO when a task is active": 469,
|
||||
"reactivates denied retry notices when authority is restored during %s": 7685,
|
||||
"reactivates dormant accepted-response work when the live chat authority registers": 2507,
|
||||
"reactivates failed retry notices from real admission after idle through %s": 3695,
|
||||
"reactivates restored ordinary input after an earlier Retry notice wake is consumed": 551,
|
||||
"reads the entire publication batch without side effects and rejects foreign bindings": 594,
|
||||
"reauthorizes a failed chat retry against a %s provider correction before lifecycle projection": 3205,
|
||||
"reauthorizes a failed chat retry and denies %s before scheduling": 3033,
|
||||
"reauthorizes skipped native progress after %s wins before retry": 1927,
|
||||
"rechecks %s authority before queued Slack Board messages and replies": 2579,
|
||||
"rechecks a recovered response at provider dispatch after %s": 3483,
|
||||
"rechecks a Slack file receipt after a competing worker changes its retry deadline": 1017,
|
||||
"rechecks canonical content before processing an automatically recovered GitHub callback": 595,
|
||||
"rechecks linked authority and channel reach before admitting a provider-confirmed Slack task (%s)": 1052,
|
||||
"rechecks pre-provider retry evidence after staging and refuses a late execution event": 694,
|
||||
"rechecks queued Slack channel roots but exempts replies and DMs (DM=$isDM, reply=$isReply)": 1695,
|
||||
"rechecks source retirement after a queued notice's late provider receipt commits": 764,
|
||||
"rechecks the outbound link when publication commits between reaction preflight reads": 1005,
|
||||
"rechecks the recovered failure link after waiting for the provider mutation lane": 866,
|
||||
"recognizes Slack callbacks behind TLS termination without trusting forwarded headers": 283,
|
||||
"reconciled chat retry delivery ownership > does not resolve or mint a substitute owner for %s": 1432,
|
||||
"reconciled chat retry delivery ownership > keeps one exact chat delivery owner through %s reconciliation": 1644,
|
||||
"reconciled chat retry delivery ownership > retains the sole owner but refuses its deferred dispatch after %s": 1237,
|
||||
"reconciles a foreign GitHub installation lifecycle event against canonical App inventory": 583,
|
||||
"reconciles a provider-confirmed Slack starter into one task without replaying Slack": 1023,
|
||||
"reconciles concurrent opposite GitHub lifecycle callbacks from canonical App state": 1186,
|
||||
"reconciles Slack membership during configure, resume, and reconnect": 733,
|
||||
"reconciles Slack session status without replaying messages after rate limits or restart": 1033,
|
||||
"reconstructs a future accepted-response deadline without early publication": 609,
|
||||
"reconstructs inbound admission committed after the notice worker stops": 473,
|
||||
"records message edits and deletes durably and deduplicates lifecycle callbacks": 797,
|
||||
"records one content-free GitHub receipt for a disabled repository without admitting work": 579,
|
||||
"records the public Cloud proxy host for authenticated Slack callback health": 345,
|
||||
"records verified Telegram edited_message updates against the existing DM task": 858,
|
||||
"recovers a committed %s answer once without rerunning or changing review/quarantine evidence": 1606,
|
||||
"recovers a genuine unanswered %s presentation": 1503,
|
||||
"recovers a provider-confirmed Slack starter after restart but rechecks revoked reach": 309,
|
||||
"recovers accepted-response work on startup and leaves selected history quiet": 1539,
|
||||
"recovers action rows on startup, wakes after outer commit, and leaves empty queues quiet": 1672,
|
||||
"recovers an accepted Slack file receipt without uploading twice or exposing provider identifiers": 1157,
|
||||
"recovers an ambiguous Slack slash-task start only after an audited explicit retry": 561,
|
||||
"recovers an authenticated GitHub webhook after shutdown before deferred work starts": 588,
|
||||
"recovers buffered Discord messages and reactions after an unreclaimed host-pause lease expiry": 1306,
|
||||
"recovers committed terminal confirmations without polling through %s": 4354,
|
||||
"recovers stale provider sends on startup without replaying uncertain effects": 356,
|
||||
"recovers the committed authorization phase after %s without renewing its deadline": 18228,
|
||||
"recovers the exact port-bound canonical response before skipping a superseded native decision": 647,
|
||||
"recovers the same Telegram bot and historical task after token rotation": 1409,
|
||||
"recovers the visible receipt exactly once after wake acceptance contention": 7724,
|
||||
"redacts lifecycle edits and ignores reactions after channel reach is revoked": 815,
|
||||
"redacts Telegram bot tokens from provider setup errors and endpoint health": 136,
|
||||
"refills four tracked publication endpoint lanes after a scheduled budget and joins shutdown": 2443,
|
||||
"refreshes Slack tool guidance without restoring disabled catalog policies": 344,
|
||||
"refuses an otherwise eligible exact Slack retry after %s without admitting source changes": 1417,
|
||||
"refuses to create a chat endpoint while its provider experiment is off": 82,
|
||||
"refuses to enable a destination the provider no longer exposes": 322,
|
||||
"registers ordinary-notice intent before a retry-owned receipt is retired": 815,
|
||||
"registers ordinary-notice work through the retry admission transaction": 575,
|
||||
"registers provider callbacks after a live Cloud claim (ingress: %s)": 1387,
|
||||
"rehydrates a durable attachment descriptor after restart and stores the file on the issue": 450,
|
||||
"rehydrates exact chat retry authority and verifies its persisted projection after service restart": 481,
|
||||
"rehydrates one Discord Gateway owner on startup and hands off cleanly across replicas": 293,
|
||||
"rejects a recovered Slack file identity when destination reach changes during lookup": 902,
|
||||
"rejects a signed automatic GitHub callback whose original %s identity changed": 2291,
|
||||
"rejects automatic pre-ingress GitHub recovery after %s": 1418,
|
||||
"rejects caller-controlled GitHub setup secrets and arbitrary credential fields": 158,
|
||||
"rejects chat endpoints for non-invokable agents": 69,
|
||||
"rejects over-scoped GitHub Apps while tolerating unavoidable lifecycle events": 225,
|
||||
"rejects reconnect credentials for a different Slack bot without rotating secrets": 158,
|
||||
"rejects under-scoped Slack and GitHub apps before saving provider credentials": 217,
|
||||
"rejects unsupported Telegram webhook ports before provider access or secret persistence": 39,
|
||||
"releases authorization row locks before a Slack slash starter waits on provider I/O": 326,
|
||||
"releases Discord ownership when runtime construction fails": 240,
|
||||
"releases task-control authorization locks before provider I/O and quarantines stale claims": 996,
|
||||
"removes a $label endpoint without deleting its task or audit history": 1258,
|
||||
"renders safe Telegram confirmations as opaque one-use buttons and wakes the resolved task": 1587,
|
||||
"reorders rapid Slack callbacks by provider time before one conversation drain": 2018,
|
||||
"reorders reverse-arrival Telegram webhooks by provider sequence before waking the agent": 1301,
|
||||
"reorders same-second GitHub callbacks by comment id before starting the task": 1393,
|
||||
"repairs a committed reply and its recovery notice without reversing row locks": 776,
|
||||
"repairs the same GitHub App webhook on reconnect without claiming a fresh signed ping or chat round trip": 638,
|
||||
"replaces a pre-connect GitHub webhook secret without inventing a broken existing App": 302,
|
||||
"replaces Telegram queued progress with its native confirmation and rejects late working progress": 1033,
|
||||
"replaces working with one safe ownership-attention message without terminalizing the run": 1060,
|
||||
"replays a pre-link reaction into the completed DM generation that sent the message": 1102,
|
||||
"replays one pre-link Discord reaction after service reconstruction without waking the task": 889,
|
||||
"requires a successful final response for setup qualification": 1053,
|
||||
"requires at-mentions by default only in Slack channels (DM=%s)": 1108,
|
||||
"requires connection-manager authority for chat connector administration": 218,
|
||||
"requires terminal-safe native ownership for an exact chat retry: %s": 9089,
|
||||
"requires the successful setup final to consume the qualifying follow-up": 1067,
|
||||
"resamples reaction expiry after replay waits on authorization locks": 849,
|
||||
"resolves ambiguous provider effects explicitly and never lets terminal failures starve reconciliation": 857,
|
||||
"resolves an admitted GitHub private image after restart with current reach at %s": 5052,
|
||||
"resource change audit atomicity > joins concurrent identical and opposite changes into actual before/after history": 124,
|
||||
"resource change audit atomicity > keeps duplicate last-write semantics but audits only net changes": 76,
|
||||
"resource change audit atomicity > keeps the audit aligned with commit on lease loss %s": 117,
|
||||
"resource change audit atomicity > reads the actual resource state after the row lock, not a pre-lock snapshot": 123,
|
||||
"resource change audit atomicity > rolls back resource writes and suppresses events on %s failure": 103,
|
||||
"resource change audit atomicity > rolls back the entire %s resource batch without audit or events": 134,
|
||||
"resource change auditing > records exact actual changes and authenticated route actor, without resource content": 182,
|
||||
"restarts an inbound notice cursor when privacy is restored during a bounded pass": 3129,
|
||||
"restores action retry and stale-claim deadlines and parks paused or uncertain work": 583,
|
||||
"restores changed native review metadata once and signals only its committed change": 1320,
|
||||
"restores chat queue deadlines without polling blocked followers or receipt owners": 376,
|
||||
"restores inbound retry and abandoned wake deadlines and excludes paused queues": 347,
|
||||
"resumes saved GitHub input after a successful board turn: %s": 2341,
|
||||
"retains a confirmation committed while another settlement is in flight": 2156,
|
||||
"retains a contended pre-run FIFO projection and settles cancelled input without a visible notice": 447,
|
||||
"retains a failed accepted-response attempt across drain and retries after resume": 724,
|
||||
"retains a retry final when its source authorization read fails transiently": 764,
|
||||
"retains a standalone unaddressed Slack thread reply for the bounded reorder window": 792,
|
||||
"retains admission committed while an empty inbound-notice selection is returning": 570,
|
||||
"retains an accepted-response deadline that elapses during the locked recovery check": 631,
|
||||
"retains an exact chat retry through a transient scheduling failure and a lost committed receipt acknowledgement": 470,
|
||||
"retains content-free GitHub source invalidation after actor revocation and denies stale retry after regrant": 1090,
|
||||
"retains HTTP Retry under the endpoint authorization row lock: %s": 5469,
|
||||
"retains inbound notices through %s during drain without passive SQL": 1473,
|
||||
"retains one HTTP task Retry intent while saved chat identity authorization is busy": 1466,
|
||||
"retains only a source invalidation for a Telegram edit when the original actor's link is revoked": 721,
|
||||
"retains the first accepted-response wake through idle drain without passive SQL": 1478,
|
||||
"retains the service idle probe after failed $name retirement and shutdown until owner retry": 1227,
|
||||
"retires a Discord receipt after terminal failure without letting an add retry resurrect it": 790,
|
||||
"retires a pre-run FIFO notice with the exact promoted %s run even before a working update": 1624,
|
||||
"retires a stale verifying receipt before another fixture runs recovery": 938,
|
||||
"retires an already-visible pre-run FIFO notice once without publishing scheduler failures": 734,
|
||||
"retires only the exact Slack final's receipt (%s)": 2802,
|
||||
"retires only the exact Telegram final receipt through the pinned adapter (%s)": 3412,
|
||||
"retires only the removed source's pre-run FIFO notice after its surviving batch promotes": 979,
|
||||
"retires superseded credentials and clears endpoint-owned secrets on removal": 419,
|
||||
"retries a colliding company fixture prefix and creates a distinct complete fixture": 88,
|
||||
"retries a contended committed review response before transport, then publishes it and its files once": 1747,
|
||||
"retries a contended pre-run FIFO receipt before I/O and observes cancellation after the lock releases": 612,
|
||||
"retries a durable lifecycle row without treating it as a message delivery": 339,
|
||||
"retries a fully recorded never-dispatched $provider setup through HTTP: $mode": 16878,
|
||||
"retries an accepted inbound wake outbox in FIFO order without blocking other conversations": 847,
|
||||
"retries an authenticated GitHub request after isolated-directory validation failed before allocation: %s": 26911,
|
||||
"retries an exact GitHub source denied by pre-provider reviewed attestation without rewriting its failure": 801,
|
||||
"retries credential cleanup after archival made a remove partially complete": 315,
|
||||
"retries failed chat execution authorization after actual %s lock contention": 18870,
|
||||
"retries lifecycle mutation atomically and reclaims it after restart": 1216,
|
||||
"retries only the original pre-provider Telegram request after exact cleanup: $mode ($errorMessage)": 27198,
|
||||
"retries pre-transport attachment integrity failures without blocking another conversation": 1362,
|
||||
"retries saved GitHub input after a successful run without replaying that run": 1103,
|
||||
"retries saved GitHub input after an endpoint admission lock collision": 31887,
|
||||
"retries the exact failed %s chat input with a new idempotent admission": 1073,
|
||||
"retries the newer saved $provider request instead of the failed request after a budget hold: $mode": 114415,
|
||||
"retries the saved board-message turn in a connected task after environment qualification failed before provider startup": 4460,
|
||||
"retries transient GitHub endpoint lock contention before accepting a wake": 879,
|
||||
"returns a committed one-time GitHub setup secret after final lease loss when its exact ref remains current": 81,
|
||||
"returns a failed link-only interaction continuation to its GitHub thread": 970,
|
||||
"returns a retryable webhook failure when the delivery insert fails before durable receipt": 1178,
|
||||
"returns ephemeral guidance for exact Slack controls in channels without creating tasks or actions": 393,
|
||||
"returns not found rather than revealing another company's chat endpoint": 45,
|
||||
"returns two answered Slack turns to Idle in the same thread (DM=%s)": 1974,
|
||||
"reuses a pre-run FIFO notice when the successor asks before its working milestone": 622,
|
||||
"reveals a rotated GitHub secret despite a later completion-audit failure and makes retry a new rotation": 1607,
|
||||
"revokes a deleted canonical GitHub installation and recovers the same endpoint after reinstall": 3454,
|
||||
"rotates a live GitHub webhook secret fail-closed and reconnects with stored credentials": 992,
|
||||
"rotates Microsoft Teams secrets without changing bot identity and preserves the last good credentials on rejection": 474,
|
||||
"round-trips a $label question modal and durably acknowledges denied or replayed tokens": 2622,
|
||||
"safely re-arms an exact terminal GitHub delivery after an operator redelivery": 666,
|
||||
"saves Slack communication instructions with validation, company isolation, and audit attribution": 259,
|
||||
"scopes board-send idempotency to one external conversation": 1427,
|
||||
"scopes failed retry notices and retains receipt contention without retrying settled history": 917,
|
||||
"scopes Slack setup %s detection to the linked user and current test": 935,
|
||||
"scopes Telegram publication failure from $label correctly": 1631,
|
||||
"segments long Telegram output into durable FIFO publications and resumes at the failed part": 1225,
|
||||
"selects the Telegram photo boundary before provider I/O (%s)": 10607,
|
||||
"sends Telegram audio and video outside native format contracts as exact original documents": 1135,
|
||||
"separates verified webhook ingress from board identity links for every webhook provider": 256,
|
||||
"serializes a GitHub secret rotation racing reconnect and preserves the rotated secret": 1166,
|
||||
"serializes and caps an agent run's selected Discord files before final handoff": 2605,
|
||||
"serializes concurrent GitHub setup-secret requests without losing stored credentials": 1597,
|
||||
"serializes Discord Gateway health events and recovers transient guild outages": 264,
|
||||
"serializes duplicate Slack admission with an owned provider-reply settlement: %s": 1324,
|
||||
"serializes recovered Telegram maintenance behind credential rotation": 443,
|
||||
"serializes setup activation ahead of a concurrent reconnect without overwriting the newer runtime": 873,
|
||||
"settles a refused GitHub task Retry by its own gates: %s": 4098,
|
||||
"settles a retried Slack receipt reaction when the provider reports already_reacted": 480,
|
||||
"settles a shared-endpoint accepted-response backlog without retrying its own locks": 2836,
|
||||
"settles an interrupted native run once without overwriting its successor": 996,
|
||||
"settles GitHub self-update receipts without orphan retries for %s": 5038,
|
||||
"settles rejected pre-acceptance intent and permits a later authorized external message": 485,
|
||||
"settles responses on different endpoints for one principal without retrying its own identity lock: %s": 4547,
|
||||
"singleflights concurrent cold GitHub runtime initialization": 678,
|
||||
"skips a locked native progress %s while unrelated final and question publications deliver": 5320,
|
||||
"skips a pre-run FIFO candidate with a noncanonical coalesced owner id (%s)": 1300,
|
||||
"skips paused outbox heads and resumes promptly without bypassing provider backoff": 1009,
|
||||
"sleeps behind an unresolved Slack tool send and wakes when its receipt settles": 453,
|
||||
"snapshots channel workspace defaults on new tasks and preserves them on replies": 777,
|
||||
"starts real Slack conversations from bare mentions and accepts same-thread follow-ups": 1581,
|
||||
"stops a conversation drain after its lease renewal fails": 560,
|
||||
"stops a Discord Gateway on lease loss and lets a standby take over": 649,
|
||||
"stops automatic pre-ingress GitHub pagination at the epoch floor without adopting older failures": 485,
|
||||
"stops starting continuation publications during %s": 4731,
|
||||
"streams long output in bounded chunks after applying the safe external projection": 734,
|
||||
"supplements Chat SDK with verified GitHub comment edit and delete lifecycle events": 2034,
|
||||
"supplements verified Teams edit, soft-delete, and restore activities against the existing task": 1351,
|
||||
"suppresses a pending pre-run FIFO notice after %s wins": 4422,
|
||||
"suppresses a queued provider reply when destination reach is revoked before transport": 352,
|
||||
"suppresses automatic pre-ingress GitHub redelivery for any existing local receipt (%s)": 2569,
|
||||
"suppresses persisted GitHub queued, progress and completion comments on replay": 696,
|
||||
"suppresses queued and working placeholders that terminalize while paused": 1090,
|
||||
"surfaces an attachment-only Telegram ingestion failure without waking the agent": 381,
|
||||
"Teams authenticated inline-picture intake > ingests an actual pinned-parser %s picture into its exact admitted turn (%s)": 3987,
|
||||
"Teams current file authority and composed Board transfer > deduplicates the same actual acceptance activity before and after file delivery": 1406,
|
||||
"Teams current file authority and composed Board transfer > durably accepts a callback inside the consent POST before the native receipt returns": 1272,
|
||||
"Teams current file authority and composed Board transfer > freezes the original recipient at Board Send and publishes only after consent and exact-byte PUT": 2219,
|
||||
"Teams current file authority and composed Board transfer > keeps the task-link fallback with no native file intent for %s": 2952,
|
||||
"Teams current file authority and composed Board transfer > offers only exact conflict cancellation after valid Teams consent receipts disagree": 1039,
|
||||
"Teams current file authority and composed Board transfer > rechecks revoked %s authority before acceptance and before PUT": 6295,
|
||||
"Teams current file authority and composed Board transfer > rejects a retained old-runtime consent callback after normal pause and resume": 1038,
|
||||
"Teams current file authority and composed Board transfer > retries only fileInfo after a confirmed PUT and an explicitly resolved missing fileInfo receipt": 1737,
|
||||
"Teams current file authority and composed Board transfer > stages native Board file consent on the first send from a cold reconstructed service": 1086,
|
||||
"Teams exact self-unknown source continuation > accepts the exact native file after a lost consent-card receipt without reposting": 1468,
|
||||
"Teams exact self-unknown source continuation > does not exempt other unknown effects or changed native authority: %s": 8488,
|
||||
"Teams exact self-unknown source continuation > retries only the confirmed native file-info stage with an exact audited phase/version, never its PUT": 2191,
|
||||
"Teams file receipt API projection > does not consume a worker slot for a pending Teams transfer on a %s endpoint": 1968,
|
||||
"Teams file receipt API projection > does not treat a caller-only transfer hint as ordinary delivery authority": 134,
|
||||
"Teams file receipt API projection > generic publication worker ignores an existing %s Teams transfer on an active endpoint": 1660,
|
||||
"Teams file receipt API projection > keeps a cancelled head locked while a consent-card receipt waits for the file tail": 125,
|
||||
"Teams file receipt API projection > never projects a transfer from another conversation into Activity": 113,
|
||||
"Teams file receipt API projection > offers only the exact %s stage actions": 453,
|
||||
"Teams file receipt API projection > preserves all-published legacy batches without transfer metadata and excludes other comments": 113,
|
||||
"Teams file receipt API projection > projects mixed delivered, declined, expired and cancelled file outcomes without making GET effects": 111,
|
||||
"Teams file receipt API projection > refuses generic replay of a failed transfer without another provider attempt": 114,
|
||||
"Teams file receipt API projection > refuses generic resolution with %s transfer preconditions without changing receipts": 654,
|
||||
"Teams file receipt API projection > refuses terminal success or actions for %s": 496,
|
||||
"Teams file receipt API projection > requires exact endpoint, conversation, publication and Board company scope": 223,
|
||||
"Teams inline picture publication > keeps $surface picture delivery truthful for $outcome": 9370,
|
||||
"Teams personal-recipient admission and durable restart > does not invent personal recipient proof when an AAD-only legacy delivery resumes without original runtime evidence": 309,
|
||||
"Teams personal-recipient admission and durable restart > preserves actual-parser personal recipient proof across database admission and reconstructed restart": 398,
|
||||
"Teams personal-recipient admission and durable restart > redacts a rejected personal destination instead of persisting its derived recipient proof": 176,
|
||||
"Telegram callback-only native private responses > cancels a queued private callback notice after %s changes without public fallback": 1223,
|
||||
"Telegram callback-only native private responses > does not rebind a retained callback denial to another %s": 754,
|
||||
"Telegram callback-only native private responses > does not refresh the callback deadline after service runtime-readiness waits": 191,
|
||||
"Telegram callback-only native private responses > does not treat another actor's private chat as a reply destination": 207,
|
||||
"Telegram callback-only native private responses > preserves an authenticated exact-actor DM denial without native ephemeral or public fallback": 246,
|
||||
"Telegram callback-only native private responses > quarantines missing private acceptance receipts and never retries them publicly": 275,
|
||||
"Telegram callback-only native private responses > records an authenticated denial before its recipient-bound send and deduplicates the exact callback": 277,
|
||||
"Telegram callback-only native private responses > retains an exact-actor DM proof across restart (expired=%s)": 447,
|
||||
"Telegram callback-only native private responses > retains the original private callback deadline across restart (expired=%s)": 533,
|
||||
"Telegram close-owned progress retirement > does not borrow an unrelated run's interaction as progress-consumer evidence": 956,
|
||||
"Telegram close-owned progress retirement > does not guess a close edit for %s": 5402,
|
||||
"Telegram close-owned progress retirement > does not reclaim a progress lane consumed by a $state $kind prompt": 3594,
|
||||
"Telegram close-owned progress retirement > keeps close uncommitted after %s of its retirement edit": 2207,
|
||||
"Telegram close-owned progress retirement > preserves an attempted unknown %s after public operator cancellation": 1994,
|
||||
"Telegram close-owned progress retirement > preserves an authored final that wins while close is admitted": 943,
|
||||
"Telegram close-owned progress retirement > qualifies $provider close-owned progress retirement ($mode)": 7384,
|
||||
"Telegram close-owned progress retirement > retires a status-interleaved working lane": 891,
|
||||
"Telegram close-owned progress retirement > retires one working lane on close without cancelling the run or publishing its late final": 1051,
|
||||
"Telegram close-owned progress retirement > treats a $state consumer with $attempts attempts conservatively": 1799,
|
||||
"Telegram current rich inbound compatibility > keeps rich files and omissions scoped through lifecycle recovery (%s)": 3025,
|
||||
"Telegram current rich inbound compatibility > retains current rich content through the pinned parser and service (%s)": 971,
|
||||
"Telegram durable private draft Stop > automatically repairs an existing %s subscription after restart without dropping updates": 6911,
|
||||
"Telegram durable private draft Stop > commits chat connection credentials atomically before waking maintenance": 930,
|
||||
"Telegram durable private draft Stop > does not confirm a subscription from non-boolean ok=%s": 1869,
|
||||
"Telegram durable private draft Stop > does not reuse an old draft ID after endpoint cascade deletion and same-bot re-binding": 1875,
|
||||
"Telegram durable private draft Stop > does not withdraw an unowned or already-claimed final (%s)": 4282,
|
||||
"Telegram durable private draft Stop > durably stops during the first draft RPC without cancelling a run, and never reuses its ID after restart": 1147,
|
||||
"Telegram durable private draft Stop > keeps ordinary output when automatic subscription repair is unsafe: %s": 11506,
|
||||
"Telegram durable private draft Stop > parks pending Telegram subscription repair while its %s is inactive": 4330,
|
||||
"Telegram durable private draft Stop > preserves an uncertain final after restart and a late Stop without automatic replay": 877,
|
||||
"Telegram durable private draft Stop > recovers Telegram maintenance from committed work and becomes SQL quiet: %s": 11961,
|
||||
"Telegram durable private draft Stop > refuses a retained Stop and later final after current %s authority changes": 3535,
|
||||
"Telegram durable private draft Stop > retries an unknown subscription mutation after restart using freshly observed options, not a recovered confirmation flag": 1378,
|
||||
"Telegram durable private draft Stop > returns 503 on failed Stop commit and accepts the exact verified retry before final send": 953,
|
||||
"Telegram durable private draft Stop > sends ordinary complete output without Stop when subscription proof is %s": 3593,
|
||||
"Telegram retained zero-message admission > recovers only ordinary source identity ($state / command=$command / $source)": 4134,
|
||||
"Telegram source-bound optional media > imports exact permitted bytes or gives accurate recovery advice (%s)": 3629,
|
||||
"Telegram source-bound optional media > retains exact durable media authority across restart (%s)": 1133,
|
||||
"tracks Slack callback surfaces independently and reports public URL drift": 311,
|
||||
"tracks the actual Board Teams claim and retires recovery when its receipt commits": 2107,
|
||||
"treats D-prefixed Slack status, new, and close callbacks as DM task controls even when the SDK flag is false": 1178,
|
||||
"turns a Slack slash command into a new native thread and one Paperclip task": 404,
|
||||
"updates only the linked same-run %s failure notice when recovering an accepted response": 1446,
|
||||
"uses a live Cloud claim after chat service startup (ingress: %s)": 265,
|
||||
"uses compact single-use Telegram action tokens and filters forged, expired, and oversized callbacks": 1447,
|
||||
"uses GitHub numeric user ids for linked authority and rechecks membership on every follow-up": 683,
|
||||
"uses the Teams group-chat toggle without weakening channel resource gates": 802,
|
||||
"wakes a failed retry notice only after %s cancellation commits": 1359,
|
||||
"wakes a not-started retry notice after resumed admission is denied": 932,
|
||||
"wakes a saved inbound notice when %s restores privacy access": 1686,
|
||||
"wakes a visible pre-run FIFO retirement only after %s commits": 5577,
|
||||
"wakes accepted-response recovery from the real coordinator retry release": 793,
|
||||
"wakes accepted-response recovery when its authenticated accepted-result journal arrives late": 2568,
|
||||
"wakes an existing chat milestone when ordinary ingress reactivates its conversation": 1417,
|
||||
"wakes chat dispatch after commit and refills free endpoint slots without the maintenance loop": 1895,
|
||||
"wakes dormant accepted-response history from the real company membership restoration owner": 2573,
|
||||
"wakes dormant accepted-response recovery after agent resume": 2831,
|
||||
"wakes inbound notice retirement from a signed Slack Stop without a maintenance scan": 901,
|
||||
"wakes interaction continuations from a committed response through %s": 12965,
|
||||
"wakes parked Slack %s actions after board %s commits": 7361,
|
||||
"wakes pre-run FIFO notices from real admission after quiet and becomes quiet again": 687,
|
||||
"wakes saved native progress when %s restores eligibility": 5559,
|
||||
"wakes Slack session sync from publication commit and leaves settled action queues quiet": 1699,
|
||||
"withholds a committed one-time GitHub setup secret after final lease loss when its ref was superseded": 70,
|
||||
"yields a %s inbound notice backlog between bounded passes without losing work": 6983
|
||||
}
|
||||
}
|
||||
@@ -219,7 +219,7 @@ export async function collectMergedPrSamples({ gh, lanes, maxRuns = 3, pullLimit
|
||||
const pulls = await gh(`${api}/pulls?state=closed&base=master&sort=updated&direction=desc&per_page=${pullLimit}`);
|
||||
const merged = pulls.filter(pull => pull.merged_at).sort((a, b) => b.merged_at.localeCompare(a.merged_at));
|
||||
for (const pull of merged) {
|
||||
const pending = lanes.filter(lane => result[lane.name].samples.length < maxRuns);
|
||||
const pending = lanes.filter(lane => result[lane.name].samples.length < (lane.maxRuns ?? maxRuns));
|
||||
if (!pending.length) break;
|
||||
validatePull(pull);
|
||||
// Only measure code present in this checkout.
|
||||
@@ -249,6 +249,7 @@ export async function collectMergedPrSamples({ gh, lanes, maxRuns = 3, pullLimit
|
||||
log(`::warning::${lane.name}: skipped run ${run.id} for PR #${pull.number}: ${error.message.split("\n")[0]}`);
|
||||
continue;
|
||||
}
|
||||
if (lane.minFiles && sample.size < lane.minFiles) continue;
|
||||
result[lane.name].samples.push(sample);
|
||||
result[lane.name].sources.push({ pull: pull.number, run: run.id, attempt: run.run_attempt, sha: run.head_sha, url: run.html_url });
|
||||
}
|
||||
|
||||
@@ -0,0 +1,110 @@
|
||||
import assert from "node:assert/strict";
|
||||
import { spawnSync } from "node:child_process";
|
||||
import { mkdtempSync, readdirSync, readFileSync, rmSync, statSync, writeFileSync } from "node:fs";
|
||||
import os from "node:os";
|
||||
import path from "node:path";
|
||||
import { fileURLToPath } from "node:url";
|
||||
|
||||
export const chatSuite = "server/src/__tests__/chat-channels.integration.test.ts";
|
||||
export const chatManifest = "scripts/chat-shard-durations.json";
|
||||
const comment = "Generated by node scripts/ci/refresh-chat-shard-durations.mjs from the CI profile test records of the PR chat jobs (profile-ci label). Per collected test entry, keyed by the name that vitest list reports: the summed duration of every case that entry runs, so an it.each line carries all its cases. Median across the given runs. scripts/test-line-shard.mjs uses these weights only to place source lines on chat partitions; they never change which tests run. Missing entries take the median weight.";
|
||||
const placeholder = /%[sdifjo#%]|\$[A-Za-z_][\w.]*|\$\d+/;
|
||||
|
||||
function literalLength(name) {
|
||||
return name.split(new RegExp(placeholder, "g")).join("").length;
|
||||
}
|
||||
|
||||
export function entryMatcher(entries) {
|
||||
const patterns = entries.map((name) => ({
|
||||
name,
|
||||
literal: literalLength(name),
|
||||
regex: new RegExp(`(?:^|> )${name.split(new RegExp(placeholder, "g"))
|
||||
.map((part) => part.replace(/[.*+?^${}()|[\]\\]/g, "\\$&")).join(".*?")}$`, "s"),
|
||||
}));
|
||||
return (caseName) => {
|
||||
let best = null;
|
||||
for (const pattern of patterns) {
|
||||
if (!pattern.regex.test(caseName)) continue;
|
||||
const exact = caseName === pattern.name || caseName.endsWith(` > ${pattern.name}`);
|
||||
const score = [exact ? 1 : 0, pattern.literal];
|
||||
if (!best || score[0] > best.score[0] || (score[0] === best.score[0] && score[1] > best.score[1])) {
|
||||
best = { name: pattern.name, score };
|
||||
}
|
||||
}
|
||||
return best?.name ?? null;
|
||||
};
|
||||
}
|
||||
|
||||
export function entryDurations(entries, cases) {
|
||||
const match = entryMatcher(entries);
|
||||
const totals = new Map();
|
||||
const unmatched = [];
|
||||
for (const { name, duration } of cases) {
|
||||
const entry = match(name);
|
||||
if (entry === null) unmatched.push(name);
|
||||
else totals.set(entry, (totals.get(entry) ?? 0) + duration);
|
||||
}
|
||||
return { totals, unmatched };
|
||||
}
|
||||
|
||||
function profileCases(directory) {
|
||||
const cases = [];
|
||||
const walk = (dir) => {
|
||||
for (const entry of readdirSync(dir)) {
|
||||
const file = path.join(dir, entry);
|
||||
if (statSync(file).isDirectory()) walk(file);
|
||||
else if (/^vitest-.+\.ndjson$/.test(entry)) {
|
||||
for (const line of readFileSync(file, "utf8").split("\n")) {
|
||||
if (!line.trim()) continue;
|
||||
const record = JSON.parse(line);
|
||||
if (record.kind !== "test" || record.file !== chatSuite) continue;
|
||||
assert.equal(typeof record.name, "string");
|
||||
assert.ok(Number.isFinite(record.duration) && record.duration >= 0 && record.duration <= 3_600_000);
|
||||
if (record.status === "passed") cases.push({ name: record.name, duration: record.duration });
|
||||
}
|
||||
}
|
||||
}
|
||||
};
|
||||
walk(directory);
|
||||
return cases;
|
||||
}
|
||||
|
||||
function collectedEntries(repoRoot) {
|
||||
const directory = mkdtempSync(path.join(os.tmpdir(), "chat-durations-"));
|
||||
try {
|
||||
const output = path.join(directory, "list.json");
|
||||
const result = spawnSync("pnpm", ["exec", "vitest", "list", "--exclude", "**/dist/**", "--project", "@paperclipai/server",
|
||||
chatSuite, "--allowOnly=false", "--includeTaskLocation", `--json=${output}`], { cwd: repoRoot, stdio: "inherit" });
|
||||
assert.equal(result.status, 0, "vitest list failed");
|
||||
return [...new Set(JSON.parse(readFileSync(output, "utf8")).map((test) => test.name))];
|
||||
} finally {
|
||||
rmSync(directory, { recursive: true, force: true });
|
||||
}
|
||||
}
|
||||
|
||||
export function chatManifestData(entries, runs, sources) {
|
||||
const samples = new Map();
|
||||
for (const cases of runs) {
|
||||
const { totals, unmatched } = entryDurations(entries, cases);
|
||||
assert.equal(unmatched.length, 0, `Cases without a collected entry: ${unmatched.slice(0, 5).join(", ")}`);
|
||||
for (const [entry, ms] of totals) samples.set(entry, [...(samples.get(entry) ?? []), ms]);
|
||||
}
|
||||
const durations = {};
|
||||
for (const [entry, values] of [...samples].sort(([a], [b]) => a.localeCompare(b))) {
|
||||
const sorted = values.sort((a, b) => a - b);
|
||||
const middle = Math.floor(sorted.length / 2);
|
||||
durations[entry] = Math.max(1, Math.round(sorted.length % 2 ? sorted[middle] : (sorted[middle - 1] + sorted[middle]) / 2));
|
||||
}
|
||||
return { $comment: comment, unit: "ms", sources, coverage: { measured: Object.keys(durations).length, entries: entries.length }, durations };
|
||||
}
|
||||
|
||||
if (process.argv[1] && path.resolve(process.argv[1]) === fileURLToPath(import.meta.url)) {
|
||||
const args = process.argv.slice(2);
|
||||
const sources = args.filter((arg) => arg.startsWith("--source=")).map((arg) => arg.slice("--source=".length));
|
||||
const directories = args.filter((arg) => !arg.startsWith("--"));
|
||||
assert.ok(directories.length > 0, "Usage: refresh-chat-shard-durations.mjs [--source=<run url>] <profile artifact directory>...");
|
||||
const repoRoot = path.resolve(path.dirname(fileURLToPath(import.meta.url)), "..", "..");
|
||||
const data = chatManifestData(collectedEntries(repoRoot), directories.map(profileCases), sources);
|
||||
writeFileSync(path.join(repoRoot, chatManifest), JSON.stringify(data, null, 2) + "\n");
|
||||
console.log(`${chatManifest}: ${data.coverage.measured} of ${data.coverage.entries} entries measured`);
|
||||
}
|
||||
@@ -16,6 +16,7 @@ const api = `/repos/${repository}`;
|
||||
const github = (endpoint, json = true) => ghApi(endpoint, { json });
|
||||
const masterComment = "Generated by node scripts/ci/refresh-test-durations.mjs from successful protected-master Cloud readiness runs. Median per-file cost includes tests, hooks, collection, setup and environment preparation when the timing reporter is present; older logs bootstrap test/hook durations only. Missing files retain their prior weight or the partitioner's median fallback. Timing changes only partition work; they never authorize skipping a test.";
|
||||
const e2eComment = "Generated by node scripts/ci/refresh-test-durations.mjs from successful .github/workflows/pr.yml runs of PRs merged into master, each pinned to the PR's final head SHA. Per spec: the sum of gaps between consecutive Playwright result lines in 'Run e2e tests' (one worker), so fixture and navigation time count; tests located in a *.shared.ts file are attributed to the shard's matching spec. Median across runs. scripts/e2e-shard.mjs uses these weights only to place specs on shards; they never change which specs run. Missing specs retain their prior weight or the median fallback.";
|
||||
const generalComment = "Generated by node scripts/ci/refresh-test-durations.mjs from successful .github/workflows/pr.yml runs of PRs merged into master, each pinned to the PR's final head SHA and each covering at least 90% of the general server lane. Per file: the cost the timing reporter prints in the General tests (server) jobs, which includes tests, hooks, collection, setup and environment preparation under the lane's own file concurrency. Median across runs. Missing files retain their prior weight or the partitioner's median fallback. Timing changes only partition work; they never authorize skipping a test.";
|
||||
const runnerComment = "Generated by node scripts/ci/refresh-test-durations.mjs from successful .github/workflows/pr.yml runs of PRs merged into master, each pinned to the PR's final head SHA. Per file: the Vitest file duration printed by the Verify Paperclip Runner package job (tests and hooks; collection and setup are not included). Median across runs. Weights only place files on lanes; they never change which tests run. Missing files retain their prior weight or the median fallback.";
|
||||
|
||||
export function validateRun(run) {
|
||||
@@ -130,6 +131,18 @@ export function parseArgs(argv) {
|
||||
return options;
|
||||
}
|
||||
|
||||
export function generalServerShardCount(matrixFile = ".github/ci-task-matrix.json") {
|
||||
const entries = JSON.parse(readFileSync(matrixFile, "utf8")).general_tests
|
||||
.filter(entry => entry.group === "general-server-without-chat");
|
||||
assert.ok(entries.length > 0 && entries.every(entry => entry.shard_count === entries.length));
|
||||
return entries.length;
|
||||
}
|
||||
|
||||
export function generalServerJob(name, fullCount = generalServerShardCount()) {
|
||||
const match = /^ci \/ General tests \(server \((\d+)\/(\d+)\)\)$/.exec(name);
|
||||
return match && Number(match[2]) === fullCount ? { index: Number(match[1]), count: Number(match[2]) } : null;
|
||||
}
|
||||
|
||||
export const serverLane = name => /General tests \(server \(\d+\/\d+\)\)$/.test(name) ? "general"
|
||||
: /Serialized tests \(\d+\/\d+\)$/.test(name) ? "serialized" : null;
|
||||
|
||||
@@ -141,7 +154,6 @@ async function refreshMasterLanes(runIds) {
|
||||
encoding: "utf8", env: { ...process.env, GITHUB_WORKFLOW: "PR", PAPERCLIP_CI_TEST_SELECTION: "" },
|
||||
}));
|
||||
const lanes = {
|
||||
general: { files: new Set(inventory.selectedGeneralServerSuites), path: "scripts/general-server-shard-durations.json", samples: [] },
|
||||
serialized: { files: new Set(inventory.selectedSerializedSuites), path: "scripts/serialized-shard-durations.json", samples: [] },
|
||||
};
|
||||
const runJobs = async run => {
|
||||
@@ -173,14 +185,14 @@ async function refreshMasterLanes(runIds) {
|
||||
if (!isMergedIntoHead(run.head_sha)) continue;
|
||||
validateRun(run);
|
||||
const jobs = await runJobs(run);
|
||||
if (jobs.some(job => serverLane(job.name))) candidates.push({ run, jobs });
|
||||
if (jobs.some(job => serverLane(job.name) === "serialized")) candidates.push({ run, jobs });
|
||||
}
|
||||
}
|
||||
assert.ok(candidates.length > 0, "Expected one to three successful master runs");
|
||||
const sources = [];
|
||||
for (const { run, jobs } of candidates) {
|
||||
const samples = { general: new Map(), serialized: new Map() };
|
||||
const selected = jobs.map(job => ({ job, lane: serverLane(job.name) })).filter(item => item.lane);
|
||||
const samples = { serialized: new Map() };
|
||||
const selected = jobs.map(job => ({ job, lane: serverLane(job.name) })).filter(item => item.lane === "serialized");
|
||||
assert.ok(selected.length, "No server timing jobs found");
|
||||
// Keep API traffic and log buffers bounded. Logs are parsed data, never code.
|
||||
for (let start = 0; start < selected.length; start += 3) {
|
||||
@@ -206,9 +218,22 @@ async function refreshMasterLanes(runIds) {
|
||||
|
||||
// Fewer than two valid runs leaves a manifest unchanged with a visible warning;
|
||||
// one run is too noisy to replace measured weights.
|
||||
export function generalServerInventory() {
|
||||
return new Set(JSON.parse(execFileSync(process.execPath, ["scripts/run-vitest-stable.mjs", "--mode", "general",
|
||||
"--group", "general-server", "--shard-index", "0", "--shard-count", "1", "--dry-run"], {
|
||||
encoding: "utf8", env: { ...process.env, GITHUB_WORKFLOW: "PR", PAPERCLIP_CI_TEST_SELECTION: "" },
|
||||
})).selectedGeneralServerSuites);
|
||||
}
|
||||
|
||||
async function refreshPrLanes(names) {
|
||||
if (!names.length) return [];
|
||||
const definitions = {
|
||||
general: () => {
|
||||
const inventory = generalServerInventory();
|
||||
const fullCount = generalServerShardCount();
|
||||
return { classify: name => generalServerJob(name, fullCount), parse: parseTimings, inventory, path: "scripts/general-server-shard-durations.json",
|
||||
comment: generalComment, maxRuns: 2, minFiles: Math.ceil(inventory.size * 0.9) };
|
||||
},
|
||||
e2e: () => ({ classify: e2eShardJob, parse: parseE2eLog, inventory: new Set(listE2eSpecs()), path: e2eManifest, comment: e2eComment }),
|
||||
runner: () => ({ classify: runnerPackageJob, parse: parseRunnerLog, inventory: runnerInventory(), path: runnerManifest, comment: runnerComment }),
|
||||
};
|
||||
@@ -232,7 +257,8 @@ export async function refreshDurations(argv = []) {
|
||||
const options = parseArgs(argv.map(String));
|
||||
const updates = [
|
||||
...(options.lanes.has("server") ? await refreshMasterLanes(options.runIds) : []),
|
||||
...await refreshPrLanes(["e2e", "runner"].filter(lane => options.lanes.has(lane))),
|
||||
...await refreshPrLanes([...(options.lanes.has("server") ? ["general"] : []),
|
||||
...["e2e", "runner"].filter(lane => options.lanes.has(lane))]),
|
||||
];
|
||||
// Validate every lane before replacing any checked-in manifest.
|
||||
for (const { path: file, previous, data } of updates) {
|
||||
|
||||
@@ -6,7 +6,8 @@ import OversizedSpecReporter from "./oversized-test-files-playwright.mjs";
|
||||
import OversizedTestFileReporter, { oversizedAnnotation, oversizedThresholdMs } from "./oversized-test-files.mjs";
|
||||
import { assertManifestSource, attributeSpec, collectMergedPrSamples, completeShardSet, e2eShardJob, parseE2eLog, parseRunnerLog,
|
||||
runnerInventory, runnerPackageJob, validatePrRun, validatePull } from "./pr-run-durations.mjs";
|
||||
import { oversizedFiles, oversizedPlaceholder, parseArgs, parseTimings, renderPrBody, serverLane, updatedDurations } from "./refresh-test-durations.mjs";
|
||||
import { generalServerJob, generalServerShardCount, oversizedFiles, oversizedPlaceholder, parseArgs, parseTimings, renderPrBody, serverLane,
|
||||
updatedDurations } from "./refresh-test-durations.mjs";
|
||||
import { timingPrefix } from "./test-timing-reporter.mjs";
|
||||
|
||||
const esc = "\u001b";
|
||||
@@ -235,6 +236,49 @@ test("collection skips unmerged, absent, stale-head and incomplete runs, and sam
|
||||
assert.ok(!requested.some(endpoint => endpoint.includes(sha("3")) || endpoint.includes(sha("4"))));
|
||||
});
|
||||
|
||||
test("general server weights come from complete full-width PR shard sets that cover the lane", async () => {
|
||||
assert.equal(generalServerShardCount(), 12);
|
||||
assert.deepEqual(generalServerJob("ci / General tests (server (3/12))", 12), { index: 3, count: 12 });
|
||||
for (const name of ["ci / General tests (server (1/3))", "ci / General tests (chat (1/3))", "ci / Serialized tests (1/9)"]) {
|
||||
assert.equal(generalServerJob(name, 12), null);
|
||||
}
|
||||
const api = "/repos/paperclipai/paperclip";
|
||||
const timing = (file, ms) => `2026-11-13T00:00:00.0000000Z ${timingPrefix}${JSON.stringify({ version: 1, file, durationMs: ms })}`;
|
||||
const files = ["server/src/a.test.ts", "server/src/b.test.ts"];
|
||||
const jobs = (run, ids) => ({ jobs: ids.map((id, index) => ({ id, run_id: run, conclusion: "success",
|
||||
name: `ci / General tests (server (${index + 1}/${ids.length}))` })) });
|
||||
const responses = {
|
||||
[`${api}/pulls?state=closed&base=master&sort=updated&direction=desc&per_page=40`]: [pull(1, sha("1"), sha("a")),
|
||||
pull(2, sha("2"), sha("b")), pull(3, sha("3"), sha("c")), pull(4, sha("4"), sha("d"))],
|
||||
...Object.fromEntries([1, 2, 3, 4].map(n => [
|
||||
`${api}/actions/workflows/pr.yml/runs?head_sha=${sha(String(n))}&event=pull_request&status=success&per_page=20`,
|
||||
{ workflow_runs: [prRun(n * 10, sha(String(n)))] }])),
|
||||
[`${api}/actions/runs/10/jobs?filter=latest&per_page=100&page=1`]: jobs(10, [101, 102]),
|
||||
[`${api}/actions/runs/20/jobs?filter=latest&per_page=100&page=1`]: jobs(20, [201]),
|
||||
[`${api}/actions/runs/30/jobs?filter=latest&per_page=100&page=1`]: jobs(30, [301, 302]),
|
||||
[`${api}/actions/runs/40/jobs?filter=latest&per_page=100&page=1`]: jobs(40, [401, 402]),
|
||||
[`${api}/actions/jobs/101/logs`]: timing(files[0], 1000),
|
||||
[`${api}/actions/jobs/102/logs`]: timing(files[1], 3000),
|
||||
[`${api}/actions/jobs/301/logs`]: timing(files[0], 2000),
|
||||
[`${api}/actions/jobs/302/logs`]: "",
|
||||
[`${api}/actions/jobs/401/logs`]: timing(files[0], 4000),
|
||||
[`${api}/actions/jobs/402/logs`]: timing(files[1], 5000),
|
||||
};
|
||||
const requested = [];
|
||||
const gh = async endpoint => {
|
||||
requested.push(endpoint);
|
||||
assert.ok(endpoint in responses, `unexpected request ${endpoint}`);
|
||||
return responses[endpoint];
|
||||
};
|
||||
const lane = { name: "general", classify: name => generalServerJob(name, 2), parse: parseTimings,
|
||||
inventory: new Set(files), maxRuns: 2, minFiles: 2 };
|
||||
const result = await collectMergedPrSamples({ gh, lanes: [lane], isMerged: () => true, log: () => {} });
|
||||
assert.deepEqual(result.general.sources.map(source => source.run), [40, 10]);
|
||||
assert.deepEqual(result.general.samples.map(sample => Object.fromEntries([...sample].map(([file, value]) => [file, value.ms]))),
|
||||
[{ [files[0]]: 4000, [files[1]]: 5000 }, { [files[0]]: 1000, [files[1]]: 3000 }]);
|
||||
assert.ok(!requested.some(endpoint => endpoint.includes("/jobs/201/")), "a narrower selected run is never downloaded");
|
||||
});
|
||||
|
||||
test("merged-PR manifests use the shared schema and only reorder placement", () => {
|
||||
const inventory = new Set(["tests/e2e/a.spec.ts", "tests/e2e/b.spec.ts"]);
|
||||
const samples = [new Map([["tests/e2e/a.spec.ts", { ms: 1000, complete: true }], ["tests/e2e/b.spec.ts", { ms: 50, complete: true }]]),
|
||||
|
||||
+1788
-1723
File diff suppressed because it is too large.
Load diff
@@ -1,5 +1,5 @@
|
||||
#!/usr/bin/env node
|
||||
import { spawn, spawnSync } from "node:child_process";
|
||||
import { spawn } from "node:child_process";
|
||||
import { mkdirSync, mkdtempSync, readdirSync, readFileSync, realpathSync, statSync, writeFileSync } from "node:fs";
|
||||
import os from "node:os";
|
||||
import path from "node:path";
|
||||
@@ -19,6 +19,7 @@ const generalServerShardDurations = loadShardDurations(
|
||||
const serializedShardDurations = loadShardDurations(
|
||||
path.join(scriptsDir, "serialized-shard-durations.json"),
|
||||
);
|
||||
const chatShardDurations = loadShardDurations(path.join(scriptsDir, "chat-shard-durations.json"));
|
||||
const serverRoot = path.join(repoRoot, "server");
|
||||
const serverSrcDir = path.join(repoRoot, "server", "src");
|
||||
const serverTestsDir = path.join(repoRoot, "server", "src", "__tests__");
|
||||
@@ -409,8 +410,7 @@ function selectSerializedSuites(routeTests, shardIndex, shardCount) {
|
||||
return shardFiles.map((file) => byRepoPath.get(file));
|
||||
}
|
||||
|
||||
async function runVitest(args, label, testShard = null, exactFiles = false) {
|
||||
console.log(`\n[test:run] ${label}`);
|
||||
function isolatedTestEnv() {
|
||||
invocationIndex += 1;
|
||||
const tempRootParent = process.platform === "win32" ? os.tmpdir() : "/tmp";
|
||||
// Production workspace/security checks reject symlink aliases. In particular
|
||||
@@ -435,19 +435,33 @@ async function runVitest(args, label, testShard = null, exactFiles = false) {
|
||||
delete env.PAPERCLIP_CI_SELECTION_HASH;
|
||||
mkdirSync(env.PAPERCLIP_HOME, { recursive: true });
|
||||
mkdirSync(env.TMPDIR, { recursive: true });
|
||||
return { testRoot, env };
|
||||
}
|
||||
|
||||
async function collectVitestList(filters, { testRoot, env }, name) {
|
||||
const output = path.join(testRoot, `${name}.json`);
|
||||
await new Promise((resolve, reject) => {
|
||||
const child = spawn("pnpm", ["exec", "vitest", "list", ...sourceOnlyVitestArgs,
|
||||
...filters, "--allowOnly=false", "--includeTaskLocation", `--json=${output}`], {
|
||||
cwd: repoRoot, env, stdio: "inherit",
|
||||
});
|
||||
child.once("error", (error) => reject(new Error(`Vitest collection failed: ${error.message}`)));
|
||||
child.once("close", (code, signal) => {
|
||||
if (code === 0) resolve();
|
||||
else reject(new Error(`Vitest collection failed: ${signal ?? code}`));
|
||||
});
|
||||
});
|
||||
return JSON.parse(readFileSync(output, "utf8"));
|
||||
}
|
||||
|
||||
async function runVitest(args, label, testShard = null, exactFiles = false) {
|
||||
console.log(`\n[test:run] ${label}`);
|
||||
const isolated = isolatedTestEnv();
|
||||
const { env } = isolated;
|
||||
if (testShard) {
|
||||
const file = path.resolve(repoRoot, chatSuite);
|
||||
const sourceLines = readFileSync(file, "utf8").split("\n");
|
||||
const collect = (filters, name) => {
|
||||
const output = path.join(testRoot, `${name}.json`);
|
||||
const result = spawnSync("pnpm", ["exec", "vitest", "list", ...sourceOnlyVitestArgs,
|
||||
...filters, "--allowOnly=false", "--includeTaskLocation", `--json=${output}`], {
|
||||
cwd: repoRoot, env, stdio: "inherit",
|
||||
});
|
||||
if (result.error || result.status !== 0) throw new Error(`Vitest collection failed: ${result.error?.message ?? result.status}`);
|
||||
return JSON.parse(readFileSync(output, "utf8"));
|
||||
};
|
||||
const allCollected = collect(args, "all");
|
||||
const allCollected = testShard.collected ?? await collectVitestList(args, isolated, "all");
|
||||
const collected = [];
|
||||
const bogusCollected = [];
|
||||
for (const test of allCollected) {
|
||||
@@ -458,15 +472,15 @@ async function runVitest(args, label, testShard = null, exactFiles = false) {
|
||||
`[test:run] dropping ${bogusCollected.length} bogus Vitest collection entr${bogusCollected.length === 1 ? "y" : "ies"} with no matching it()/test() call: ${bogusCollected.map((test) => `${test.location.line}:${JSON.stringify(test.name)}`).join(", ")}`,
|
||||
);
|
||||
}
|
||||
const selected = partitionTestLines(collected, testShard.count, file)[testShard.index];
|
||||
const selected = partitionTestLines(collected, testShard.count, file, chatShardDurations)[testShard.index];
|
||||
// Validate shard membership with the same `list`-reported lines that
|
||||
// produced `selected`, then switch to the `run`-compatible lines only for
|
||||
// the filters handed to the real `vitest run` below (see
|
||||
// chatSuiteRunLine).
|
||||
const validationFilters = selected.lines.map((line) => `${chatSuite}:${line}`);
|
||||
const validationArgs = [...args.filter((arg) => arg !== chatSuite), ...validationFilters];
|
||||
assertSelectedTests(selected.tests, collect(validationArgs, "selected"), file);
|
||||
console.log(`[test:run] chat shard ${testShard.index + 1}/${testShard.count}: ${selected.tests.length}/${collected.length} tests, ${selected.lines.length} source lines; exact filter coverage verified`);
|
||||
assertSelectedTests(selected.tests, await collectVitestList(validationArgs, isolated, "selected"), file);
|
||||
console.log(`[test:run] chat shard ${testShard.index + 1}/${testShard.count}: ${selected.tests.length}/${collected.length} tests, ${selected.lines.length} source lines, ${Math.round(selected.weight / 1000)}s recorded; exact filter coverage verified`);
|
||||
const runFilters = selected.lines.map((line) => `${chatSuite}:${chatSuiteRunLine(sourceLines, line)}`);
|
||||
args = [...args.filter((arg) => arg !== chatSuite), ...runFilters, "--allowOnly=false"];
|
||||
}
|
||||
@@ -511,7 +525,7 @@ async function runProjectGroup(projects, groupName, shardIndex = null, shardCoun
|
||||
}
|
||||
|
||||
function chatShardWorkers(index, count) {
|
||||
const workers = count > 1 && !hasSharedTestDatabase && !options.serialChat ? 2 : 1;
|
||||
const workers = count > 1 && !hasSharedTestDatabase && !options.serialChat ? 3 : 1;
|
||||
return Array.from({ length: workers }, (_, offset) => ({
|
||||
index: (index ?? 0) * workers + offset,
|
||||
count: (count ?? 1) * workers,
|
||||
@@ -522,9 +536,11 @@ async function runGeneralGroup(routeTests, groupName, shardIndex = null, shardCo
|
||||
if (groupName === generalChatGroupName) {
|
||||
// Each invocation gets its own temporary home and embedded PostgreSQL.
|
||||
// Wait for both children even on failure so no sibling is abandoned.
|
||||
const results = await Promise.allSettled(chatShardWorkers(shardIndex, shardCount).map((shard) =>
|
||||
runVitest(["--project", "@paperclipai/server", ...serializedServerVitestArgs, chatSuite],
|
||||
"chat integration test shard", shard),
|
||||
const args = ["--project", "@paperclipai/server", ...serializedServerVitestArgs, chatSuite];
|
||||
const workers = chatShardWorkers(shardIndex, shardCount);
|
||||
const collected = await collectVitestList(args, isolatedTestEnv(), "all");
|
||||
const results = await Promise.allSettled(workers.map((shard) =>
|
||||
runVitest(args, "chat integration test shard", { ...shard, collected }),
|
||||
));
|
||||
const failed = results.find((result) => result.status === "rejected");
|
||||
if (failed) throw failed.reason;
|
||||
|
||||
@@ -12,23 +12,32 @@ function validateTests(tests, file) {
|
||||
}
|
||||
|
||||
// Keep all cases registered on one source line together, including it.each
|
||||
// and loop-generated cases. Balance by collected case count, not line count.
|
||||
export function partitionTestLines(tests, count, file) {
|
||||
// and loop-generated cases.
|
||||
export function partitionTestLines(tests, count, file, durations = {}) {
|
||||
validateTests(tests, file);
|
||||
assert.ok(Number.isSafeInteger(count) && count > 0, "invalid shard count");
|
||||
const recorded = Object.values(durations).filter((ms) => Number.isFinite(ms) && ms >= 0).sort((a, b) => a - b);
|
||||
const median = recorded.length === 0 ? 1
|
||||
: recorded.length % 2 ? recorded[(recorded.length - 1) / 2]
|
||||
: (recorded[recorded.length / 2 - 1] + recorded[recorded.length / 2]) / 2;
|
||||
const weightOf = (test) => recorded.length === 0 ? 1
|
||||
: Number.isFinite(durations[test.name]) && durations[test.name] >= 0 ? durations[test.name] : median;
|
||||
const byLine = new Map();
|
||||
for (const test of tests) {
|
||||
const line = test.location.line;
|
||||
if (!byLine.has(line)) byLine.set(line, []);
|
||||
byLine.get(line).push(test);
|
||||
if (!byLine.has(line)) byLine.set(line, { cases: [], weight: 0 });
|
||||
const group = byLine.get(line);
|
||||
group.cases.push(test);
|
||||
group.weight += weightOf(test);
|
||||
}
|
||||
assert.ok(byLine.size >= count, "each shard must contain a source line");
|
||||
const groups = [...byLine].sort((a, b) => b[1].length - a[1].length || a[0] - b[0]);
|
||||
const shards = Array.from({ length: count }, () => ({ lines: [], tests: [] }));
|
||||
for (const [line, cases] of groups) {
|
||||
const shard = shards.reduce((best, next) => next.tests.length < best.tests.length ? next : best);
|
||||
const groups = [...byLine].sort((a, b) => b[1].weight - a[1].weight || a[0] - b[0]);
|
||||
const shards = Array.from({ length: count }, () => ({ lines: [], tests: [], weight: 0 }));
|
||||
for (const [line, group] of groups) {
|
||||
const shard = shards.reduce((best, next) => next.weight < best.weight ? next : best);
|
||||
shard.lines.push(line);
|
||||
shard.tests.push(...cases);
|
||||
shard.tests.push(...group.cases);
|
||||
shard.weight += group.weight;
|
||||
}
|
||||
for (const shard of shards) shard.lines.sort((a, b) => a - b);
|
||||
return shards;
|
||||
|
||||
@@ -28,6 +28,8 @@ it("runs every active nested/parameterized fixture case exactly once through the
|
||||
afterEach(() => { active = false; });
|
||||
function record(id) { expect(active).toBe(true); appendFileSync(${JSON.stringify(trace)}, JSON.stringify(id) + "\\n"); }
|
||||
it("top-level", () => record("top"));
|
||||
it("second top-level", () => record("top-second"));
|
||||
it("third top-level", () => record("top-third"));
|
||||
describe("nested", () => {
|
||||
it("first", () => record("nested-first"));
|
||||
it("second", () => record("nested-second"));
|
||||
@@ -45,7 +47,7 @@ it("runs every active nested/parameterized fixture case exactly once through the
|
||||
expect(result.stdout).toContain("exact filter coverage verified");
|
||||
}
|
||||
const executed = readFileSync(trace, "utf8").trim().split("\n").map((line) => JSON.parse(line));
|
||||
expect(executed.sort()).toEqual(["a", "b", "c", "d", "nested-first", "nested-second", "top"]);
|
||||
expect(executed.sort()).toEqual(["a", "b", "c", "d", "nested-first", "nested-second", "top", "top-second", "top-third"]);
|
||||
|
||||
// A real assertion failure must still fail the wrapper after successful
|
||||
// collection and filter validation.
|
||||
|
||||
Reference in new issue
Block a user