Files
PaperClipAI/tests/runner-e2e/context-integrity-scoring.ts
DottaandPaperclip b17019e14d fix(agents): reduce default instructions and qualify stock harnesses (#14948)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work.
> - Its adapters supply task context and access to Paperclip skills and
tools.
> - The default hire manual and shared prompts also repeat general work
procedures.
> - Those procedures overlap with stock provider instructions and the
Paperclip skill.
> - Existing E2E fixtures supply a QA manual, so they do not qualify the
production default.
> - This pull request reduces the generic instructions and adds real
default-hire coverage.
> - The benefit is less competing guidance, with inspectable evidence
for preserved skills and task context.

## Linked Issues or Issue Description

Refs: #14920. That merged change preserves native Codex base
instructions. This PR covers the default manual, shared legacy prompts,
operational skill guidance, and the narrowly approved ACP
skill-discovery/session-environment repair for measured delivery and
credential-persistence failures.

**What existing behavior does this improve?**

New non-CEO hires without a custom bundle and legacy task/chat startup
and continuation prompts.

**Current behavior**

The shipped default manual contains 602 words. Generic task/chat prompts
and ordinary resume deltas repeat work procedures already available
through the harness and Paperclip skill.

**Proposed behavior**

The default manual contains only the eight-word company identity. Shared
startup prompts retain identity and connection guidance. Ordinary resume
deltas retain current work context without the generic execution
contract.

**Reason and benefit**

Let the stock harness guide general work. Keep Paperclip-specific
capabilities and independently test default hires, skills, ordered
comments, and chat restart.

**Breaking changes**

New default hires receive less guidance. Existing saved manuals,
explicit custom bundles, CEO templates, and specialized wake contracts
retain their behavior. The obsolete includeExecutionContract option
remains accepted for source compatibility.

## What Changed

- Reduce the default hire manual to one sentence.
- Reduce shared task/chat defaults and remove the generic
ordinary-resume contract.
- Keep connection guidance, auth, skills, custom prompts, and
specialized wake context.
- Add credential-free instruction-boundary gates and 26 explicit Product
E2E cells across eight legacy/native profiles, including two focused
Paperclip-storage cases.
- Capture public hire receipts before providers run, then grade
delivered prompts and independent task/chat outcomes.
- Add an early legacy skill API recipe for saving a task document,
checking the saved revision receipt and linking the document. Improve
stock task/heartbeat skill-selection metadata and show a clickable
Markdown UI-link example. Keep native tool completion separate.
- Advertise bounded routing descriptions and exact successfully staged
SKILL.md paths in legacy ACP Claude; keep full bodies on demand and
preserve remote path rebasing.
- Remove only the provider environment from copied persisted ACP session
records, while loading current run credentials and preserving all other
options/conversation state.
- Regenerate both capability metadata inventories and reject stale
manifests/inventories before provider admission.
- Publish the original reduction and focused skill-repair comparisons,
preserving all failures, automatic recovery, cost coverage and
limitations.

## Verification

**Behavioral qualification remains pending.** Original legacy ACP Claude
loses the issue document only in the reduced cohort beneath an unchanged
credential failure. A source-backed diagnosis finds that neither
ordinary assignment reads the staged operational skill, while the
runtime persists provider environment in session state. The new common
repairs expose skill metadata/path and omit persisted env; strict
document and credential guards stay intact. [Inspectable diagnosis and
retained
hashes](https://github.com/paperclipai/paperclip/blob/9f654db4541d3d002769c988f6e51fc0b08dadbd/doc/plans/2026-10-03-legacy-acp-claude-readiness.md).

Current repair head `de0965984ff3edf611ae6d0e7ca5c7d5ae3947bb`
incorporates master `569c7203aa24b95440682983ce7940ba1d4247bd` (merged
#14961/#15007). All 222 affected adapter tests, adapter-utils/E2E
typechecks, and final 96 variant/grader/retry calibrations pass. The
frozen historical comparator is
`c25697f4260b6f3adfea143c3ae9932e2f42986d`: 8,280 of 8,291 paths
identical, exactly two production instruction paths plus nine declared
unit expectations differ. The operational skill/discovery/environment
repairs, selected model/profile/task/core grader/auth/permissions/retry
policy are identical. Both actual launcher prepare→verify admissions
pass with zero providers. [Immutable manifest and exact
receipts](https://github.com/paperclipai/paperclip/blob/9f654db4541d3d002769c988f6e51fc0b08dadbd/doc/plans/2026-10-03-legacy-acp-claude-evidence/manifest.json).

One original legacy ACP Claude cell per variant is authorized, with
enforced single campaign attempts, 12-minute deadlines and company/agent
1,000-cent hard stops; every product recovery run/cost is counted.
Actual live outcomes are pending. Current normal CI has one failed
server shard and failed aggregate verify under diagnosis; other normal
gates including typecheck/build/Rust/all eight browser shards pass.
Fresh review completed successfully; the valid historical startup/resume
masking finding was fixed with per-invocation task/chat checks and
strict complete-snapshot capture, calibrated and resolved. Prior heads,
failures and campaigns below remain historical evidence, not checks on
this repair head.

- Prior head `36aa4d81c49a1a8f6f04b1a068fae19aa901955f` is replayed on
merged hiring master `862a5758ba0e88a33232c1f1fa645e85c38a3113`. All 52
current-head checks pass with two intentional Storybook skips, including
repository typecheck/test/build and the browser shard. Fresh Greptile is
5/5 with zero unresolved review threads. Exact-head stock prerequisites
pass 599 assertions (598 TypeScript + 1 Rust), all six gates and
retained receipt verification, zero providers/source errors. Fingerprint
`a7f5a22d860a88fe20cce213c6d5e0004788f32c8363930729aea4fd740ad16d`.
Combined catalog/hiring calibrations pass 67 assertions, E2E typecheck
and 26-cell stock discovery pass. Canonical contract/inventory checks
and the later issue-derived reference calibration are retained; that
reference-only follow-up is not live-qualified by earlier frozen runs.
- Prior full repository typecheck/build passed. The complete local
Vitest run executed 14,956 tests: 14,870 passed, 83 skipped, three
timing failures. All three affected files passed unchanged narrow
reruns; original failures remain retained. Current-head CI now passes
the full general checks; the original local failures remain retained.
- The original 24-pair default-manual/shared-prompt comparison has two
new overall classic Claude/OpenCode document-delivery failures plus an
additional legacy ACP Claude document loss beneath an unchanged
credential-guard failure (not closed by later runs), two newly passing
OpenCode ordered cases, seven unchanged failures and 13 unchanged
passes. Equal 15/24 totals do not establish behavioral equivalence.
[Complete original
report](https://github.com/paperclipai/paperclip/blob/875f4c397d9e8c3f12f39dedd59abaf1eaf5236e/doc/plans/2026-10-02-stock-harness-live-comparison.md).
- The skill-only repair holds the eight-word manual/shared prompts and
merged #14920 fixed. All four matched profile configurations and 203
fixture/behavior files match. Candidate
`abd0b628ca642c09a54a4edc56a5227402f6686e` varies only the two skill
sources against baseline `bc83fe030234439ac51279502a28803958963e2e`.
[Candidate
workflow](https://github.com/paperclipai/paperclip/actions/runs/37060885547)
and [baseline
workflow](https://github.com/paperclipai/paperclip/actions/runs/37060888047)
each pass 571 exact-source prerequisites before providers; all eight
cells clean up successfully. Failed campaigns publish successfully and
remain failed.
- Repair pairs: Claude original Fail → Pass; Claude explicit Pass →
Pass; both OpenCode cases Fail → Fail. Explicit OpenCode's handoff
worsens beneath the unchanged failing UI-link grade: baseline gives a
clickable API URL, candidate gives a code-formatted path without an
anchor. The request's usable-link wording is narrower in the UI-only
oracle. [Complete repair report and safe
projection](https://github.com/paperclipai/paperclip/blob/875f4c397d9e8c3f12f39dedd59abaf1eaf5236e/doc/plans/2026-10-02-legacy-document-skill-repair.md).
- The subsequent narrow stock metadata/link correction has two matched
Pass → Pass cases, zero new machine failures/passes and no pending
pairs. Both original-case handoff links remain deficient: candidate uses
a wrong PAP prefix, baseline supplies a bare prefix-less slug path; the
preserved original oracle only requires a durable document. Both
explicit clickable UI-link cases pass revision/content/link grading. All
four exact-source 587-check gates, single assignment runs and cleanup
pass. This does not establish fix causality because baseline also
succeeds. [Candidate
workflow](https://github.com/paperclipai/paperclip/actions/runs/37069547401)
freezes `fe9dc1e3c518825242ed889ab9c8352986f8c2ed`; [matched
baseline](https://github.com/paperclipai/paperclip/actions/runs/37069552374)
freezes `0d7ecfa96d72fba79b7f0a25052b42c0686c0488`. This is a skill-only
comparison with reduced manuals/shared prompts held constant, not a
repeat of the historical-manual comparison. Only original and clarified
explicit classic OpenCode cases are selected, two per variant/four
expected turns. 8,242 other tracked files and both profile hashes match;
protected workflows admit each exact source before credentials.
[Complete qualification
report](https://github.com/paperclipai/paperclip/blob/74d0d3d945f4c52d0814b5a845ab5bd09f33cd6b/doc/plans/2026-10-02-opencode-skill-routing-link-qualification.md).
Candidate original loads Paperclip/reference before saving publicly;
baseline original loads it after writing locally, then saves publicly
within the same assignment. Reported cost totals are $0.0107824490
candidate / $0.0107909015 baseline, with unmetered runtime. The later
reference-only issue-derived link correction is provider-free calibrated
and **not live-qualified** by these frozen runs; no further paid runs.
- Retained tool calls show the repaired original OpenCode assignment
loads only its assigned output skill before writing locally. Operational
Paperclip is first loaded during automatic disposition recovery; its
early recipe is visible then, but it never saves the missing document.
Explicit candidate loads Paperclip and reads the new reference before
saving successfully. All nine actual runs are counted. Reported LLM
totals are $0.3802537209 baseline and $0.4918990161 candidate; local
runtime is unmetered.
- Initial setup, packaging, cancelled/missing-cell recovery, callback
test and relative-output attempts remain retained. No completed provider
failure was rerun. Frozen measurement branches are unchanged by later
canonical metadata maintenance.
- Run `pnpm test:e2e:runner:stock-harness`, `pnpm test:e2e:runner:unit`,
and `pnpm test:e2e:runner:typecheck`. Select `stock-harness` explicitly
for paid execution; it is excluded from `--all`.

Prior-head integration: `36aa4d81c49a1a8f6f04b1a068fae19aa901955f`
replays this PR on merged hiring #14985
(`862a5758ba0e88a33232c1f1fa645e85c38a3113`), preserving the four
explicit custom-CEO-bundle checks, minimal generic manual boundary, and
both suites. The combined fixture catalog and hiring calibrations pass
67 assertions; exact-head stock prerequisites pass 599 assertions (598
TypeScript + 1 Rust), all six gates and retained-receipt verification,
zero providers/source errors, fingerprint
`a7f5a22d860a88fe20cce213c6d5e0004788f32c8363930729aea4fd740ad16d`. E2E
typecheck and 26-cell stock discovery pass. Fresh current-head CI passes
all 52 checks with two intentional skips, and fresh Greptile is 5/5 with
zero unresolved review threads.

The prior source-plan browser failure is retained: a deterministic
process fixture replayed its last `fixture:plan` command on
`chat_task_completed`, writing revision 2 with identical body after the
approval handoff. This was not paid provider execution. Rebased
current-head CI passes the same assertion without an old-head retry or a
change to that browser fixture.

The merged hiring change was measured separately on immutable matched
unions, with this reduced/shared/operational context and native
completion guidance held constant. [Complete original two-profile
report](https://github.com/paperclipai/paperclip/blob/f0512647656be78e48abd8c22a3078db8bf6bcd2/doc/plans/2026-10-02-hiring-template-live-comparison.md):
[candidate](https://github.com/paperclipai/paperclip/actions/runs/37075466208)
/ [historical
baseline](https://github.com/paperclipai/paperclip/actions/runs/37075469463),
705 provider-free prerequisites each. Both pairs are unchanged Fail →
Fail on the exact-five count, with six core delivery checks passing all
four cells; 28 actual successful runs include eight automatic completion
wakes, zero retries, four successful cleanups. Source-read coverage is
uncomparable, actual model charges unknown. Separately versioned
provider-free accounting remains analytical work; original verdicts are
preserved. This does not rerun or qualify the completed default-manual
or native campaigns.

## Risks

- Legacy ACP Claude's additional delivery loss is not closed by any
later matched run and blocks the no-extra-failing-behavior merge
criterion. Legacy document delivery may have relied on the prior
manual/shared prompts. The early skill repair improves Claude in one
trial; the later OpenCode pairs pass in both variants and cannot
establish causality or robust recovery. Both original-case links remain
deficient beneath the storage-only grade. The later issue-derived
reference correction has only provider-free validation. Native
finish/block descriptions must not be supplied to legacy agents.
- The comparison holds merged native Codex fix #14920 constant; it
cannot measure that fix's before/after task performance.
- These bounded skill/context/chat workflows do not measure general
coding quality. Unrepresented providers remain unqualified.
- Saved manuals and old Codex sessions are not automatically migrated.
Codex through ACP still has a separate base-instruction follow-up.

## Model Used

OpenAI Codex, GPT-6 family as identified by this session. The exact
deployment ID and context-window size are not exposed. The assistant
used reasoning, repository tools, code execution, and delegated PR/eval
work.

## 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 (relevant suites and all
three unchanged narrow reruns pass; complete-run timing failures
retained in Verification)
- [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 on the new repair head
(prior-head checks retained above)
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
on the new repair head
- [x] I will address all Greptile and reviewer comments before
requesting merge

---------

Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-10-03 12:32:42 -05:00

154 lines
10 KiB
TypeScript

import { isAssignedSkillContext, PAPERCLIP_DOCUMENT_CASE, type ContextIntegrityCase } from "./context-integrity-cases.js";
export interface ContextIntegrityCheckpoint {
phase: "initial" | "comment-1" | "comment-2" | "comment-3" | "final";
issue: { id: string; status: string; identifier?: string; issuePrefix?: string; appOrigin?: string };
comments: Array<Record<string, unknown>>;
queuedComments?: Record<string, unknown>;
documents: Array<{ key: string; body?: string | null; latestRevisionId?: string | null; latestRevisionNumber?: number | null }>;
runs: Array<Record<string, unknown>>;
assignedSkill?: { key: string; runtimeName?: string; versionId?: string | null; markdown?: string };
skillRequestText?: string;
runEvents?: Array<Record<string, unknown>>;
runLogs?: unknown[];
skillInvocationEvidence?: boolean;
}
function userComments(checkpoint: ContextIntegrityCheckpoint) {
return checkpoint.comments
.filter(
(comment) =>
!comment.createdByRunId &&
(comment.authorType === "user" || comment.authorUserId),
)
.map((comment) => String(comment.body ?? ""));
}
function humanCommentRows(checkpoint: ContextIntegrityCheckpoint) {
return checkpoint.comments.filter(
(comment) =>
!comment.createdByRunId &&
(comment.authorType === "user" || comment.authorUserId),
);
}
function oneOutput(checkpoint: ContextIntegrityCheckpoint, marker: string, requireMarker: boolean) {
const outputs = checkpoint.documents.filter((document) =>
!requireMarker || String(document.body ?? "").includes(marker),
);
return outputs.length === 1 ? outputs[0] : undefined;
}
function wakeCommentIds(run: Record<string, unknown>): string[] {
const context = (run.contextSnapshot ?? {}) as Record<string, unknown>;
const paperclipWake = (context.paperclipWake ?? {}) as Record<string, unknown>;
const continuation = (context.executionContinuation ?? {}) as Record<string, unknown>;
for (const candidate of [paperclipWake.commentIds, paperclipWake.wakeCommentIds, continuation.commentIds, continuation.wakeCommentIds]) {
if (Array.isArray(candidate)) return candidate.map(String);
}
return [];
}
export function gradeContextIntegrity(input: {
id: ContextIntegrityCase;
marker: string;
comments: readonly string[];
checkpoints: readonly ContextIntegrityCheckpoint[];
}) {
const checks: Array<{ id: string; passed: boolean; detail: string }> = [];
const check = (id: string, passed: boolean, detail: string) => checks.push({ id, passed, detail });
const initial = input.checkpoints.find((checkpoint) => checkpoint.phase === "initial");
const final = input.checkpoints.find((checkpoint) => checkpoint.phase === "final");
const finalComments = final ? userComments(final) : [];
if (input.id === "ordered-comment-continuation") {
const queued = input.checkpoints.find((checkpoint) => checkpoint.phase === "comment-3")?.queuedComments;
const queuedEntries = Array.isArray(queued?.entries) ? queued.entries : [];
const queuedRows = queuedEntries.map((entry) => {
const comment = (entry as Record<string, unknown>).comment as Record<string, unknown> | undefined;
return { id: String(comment?.id ?? ""), body: String(comment?.body ?? "") };
});
check("comments-queued-as-batch", queuedRows.length >= input.comments.length && input.comments.every((body, index) => queuedRows[index]?.body === body) && queuedRows.slice(0, input.comments.length).every((row) => Boolean(row.id)) && new Set(queuedRows.slice(0, input.comments.length).map((row) => row.id)).size === input.comments.length, "The three public comments must be present in one ordered deferred queue with distinct durable IDs before the initial run closes.");
const initialUserComments = (initial?.comments ?? []).filter((comment) => !comment.createdByRunId && (comment.authorType === "user" || comment.authorUserId));
check("initial-run-started", initial?.runs.some((run) => run.status === "running") === true && initialUserComments.length === 0, "The initial public run must be running before follow-up comments are submitted, with no follow-up user comments in the initial checkpoint.");
check(
"ordered-comments",
input.comments.every((body, index) => finalComments[index] === body) && finalComments.length >= input.comments.length,
"The three user comments must remain distinct and ordered, including the intentional repeated comment.",
);
const humanRows = final ? humanCommentRows(final) : [];
const ids = humanRows.map((comment) => String(comment.id ?? ""));
check(
"distinct-comment-identities",
ids.length === new Set(ids).size && ids.every(Boolean),
"Each delivered user comment must retain its own durable identity, including repeated wording.",
);
check(
"changed-scope-preserved",
finalComments.at(2) === input.comments[2] && finalComments.at(0) === finalComments.at(1),
"The final changed-scope comment must follow two identical earlier comments.",
);
const report = final?.documents.find((document) => /passport/i.test(String(document.body ?? "")) && /charger/i.test(String(document.body ?? "")));
const body = String(report?.body ?? "");
const orderedTerms = [input.comments[0], input.comments[1], input.comments[2]];
let cursor = -1;
const ordered = orderedTerms.every((term) => {
const position = body.indexOf(term, cursor + 1);
if (position < 0) return false;
cursor = position + term.length - 1;
return true;
});
check(
"packing-report-order",
/passport/i.test(body) && /charger/i.test(body) && ordered,
"The durable packing report must retain both initial items and each verbatim request in order, including the repeated request and final scope.",
);
check(
"final-scope-applied",
["launch checklist", "the launch checklist"].includes(
body.match(/(?:^|\n)##[ \t]+Final scope[ \t]*\r?\n([\s\S]*?)(?=\n#{1,6}[ \t]+|$)/i)?.[1]?.trim().toLowerCase() ?? "",
),
"The report must apply the final Launch checklist scope in a separate final-scope section outside the quoted request ledger.",
);
const succeededRunIds = (final?.runs ?? []).filter((run) => run.status === "succeeded").map((run) => String(run.id ?? "")).filter(Boolean);
check("continuation-run-count", new Set(succeededRunIds).size === 2, "The initial run and one distinct successful deferred continuation run must both be recorded.");
const continuation = (final?.runs ?? [])
.filter((run) => run.status === "succeeded")
.sort((a, b) => Date.parse(String(a.startedAt ?? "")) - Date.parse(String(b.startedAt ?? "")))[1];
const expectedCommentIds = humanRows.slice(0, input.comments.length).map((comment) => String(comment.id ?? ""));
check("continuation-wake-comment-ids", Boolean(continuation) && JSON.stringify(wakeCommentIds(continuation)) === JSON.stringify(expectedCommentIds), "The deferred continuation wake must carry all public comment IDs in arrival order.");
}
if (isAssignedSkillContext(input.id)) {
const runtimeName = String(initial?.assignedSkill?.runtimeName ?? initial?.assignedSkill?.key ?? "");
const requestText = String(initial?.skillRequestText ?? "");
const references = requestText.split(/\s+/).map((word) => word.replace(/[.,]$/, ""));
check("assigned-skill-present", Boolean(initial?.assignedSkill?.key && initial.assignedSkill.versionId), "The task run must receive one pinned skill version through the public assignment state.");
check("skill-request-explicit", Boolean(runtimeName && references.some((word) => word === `/${runtimeName}` || word === `$${runtimeName}`)), "The task request must explicitly name the assigned skill with a supported slash or dollar reference.");
check("skill-source-marker", Boolean(initial?.assignedSkill?.markdown?.includes(input.marker)), "The assigned pinned skill source must contain the output marker.");
check("marker-not-in-request", !requestText.includes(input.marker) && !finalComments.some((comment) => comment.includes(input.marker)), "The output marker must originate from the assigned skill, not the task request or comments.");
}
const output = final ? oneOutput(final, input.marker, isAssignedSkillContext(input.id)) : undefined;
check("single-durable-output", Boolean(output) && final!.documents.length === 1, isAssignedSkillContext(input.id) ? "Exactly one durable task document must contain the skill's marker." : "Exactly one durable packing report document must be saved.");
if (input.id === PAPERCLIP_DOCUMENT_CASE) {
check("saved-document-revision", Boolean(output?.latestRevisionId) && Number.isInteger(output?.latestRevisionNumber) && Number(output?.latestRevisionNumber) > 0,
"The public saved document must have a persisted revision and the requested content.");
const route = output && final?.issue.identifier && final.issue.issuePrefix
? `/${final.issue.issuePrefix}/issues/${final.issue.identifier}#document-${output.key}` : undefined;
const agentComments = (final?.comments ?? []).filter(comment => comment.authorType === "agent" || Boolean(comment.authorAgentId));
const origin = final?.issue.appOrigin;
const usableLink = Boolean(route && origin && agentComments.some(comment => {
const links = String(comment.body ?? "").matchAll(/\[[^\]]*\]\((?:<([^>]+)>|([^\s)]+))(?:\s+["'][^"']*["'])?\)/g);
return [...links].some(link => {
try {
const url = new URL(link[1] ?? link[2]!, origin);
return url.origin === origin && `${url.pathname}${url.hash}` === route && !url.search;
} catch { return false; }
});
}));
check("saved-document-link", usableLink,
"An agent completion comment must link the exact saved document on this task; a local path or claimed URL does not suffice.");
}
check("completed-task", final?.issue.status === "done", `Final task status: ${final?.issue.status ?? "missing"}.`);
check("successful-runs", Boolean(final?.runs.length) && final!.runs.every((run) => run.status === "succeeded"), "All recorded context-integrity runs must succeed.");
return checks;
}