mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-14 08:53:19 +01:00
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work > - The chat-feature evals measure whether agents complete real chat requests through the product > - The chat catalog eval harness and its repairs live on the `chat-eval/master-candidate` branch, not on master > - Today's work made 139 of 242 chat evals pass in verified single-case live runs. Each product fix landed on master as its own pull request > - The eval-side changes (setups, graders, fault seams, evidence capture) still exist only on the candidate > - This pull request lands the harness on master. Master stays the source of truth for product behavior > - The benefit is one tree: evals run against master commits, and product fixes reach the evals as soon as they merge ## Linked Issues or Issue Description Refs #19441, Refs #19599, Refs #20270, Refs #20334, Refs #20342, Refs #20344, Refs #20354, Refs #20360, Refs #20373, Refs #20382, Refs #20409 **What existing behavior does this improve?** The chat catalog eval harness and the product it evaluates live on a long-running branch outside master. Eval fixes do not reach master, so master-based eval runs use old setups and graders. **Subsystem affected** `tests/runner-e2e` chat catalog harness, plus a small set of product changes that exist only on the candidate. **Proposed change** Land the harness on master. Every product fix from this eval work already merged to master as its own pull request. Conflicts keep master's product behavior. ## What Changed - Most of the diff (808 of 851 files) is under `tests/runner-e2e/`: chat-catalog domains, scenarios, descriptors, the TestOps manifest, private evidence tooling, and their tests. - Graders now check saved state, setup literals and rendered or clicked links. Many keyword and regex checks on free-form replies are removed. Each changed grader has a bumped `oracleVersion` and calibration tests (a valid reply, a reworded valid reply, and broken variants that must fail). - The shared link helper `hasLink` no longer treats a list link (for example `/COM/issues?view=mine`) as a link to every task without a URL key. - Setup fixes build real state through supported product routes: routine history, agent grants through the board grant route, budget holds with a real cap, and a storage-layer archive fault for `inbox-search.archive-partial-failure`. - Browser boundary journeys (sign-in, chat-disabled, onboarding) now write the standard pass evidence files. - Retargeted cases, approved by the maintainer: `agent-chat.chat-disabled` (Conference Room gate), `agents.restore-instructions` (no saved instruction history), `access.manage-membership` (member permission edits stay board-only). - The private evidence trusted files (`manual-private-handoff.ts`, `manual-private-retention.ts`, `manual-private-report-assets.ts`) are byte-identical to master. - Product files outside `tests/` (43 files, +1,328 −49) carry candidate-only changes that master does not yet have. Most are tests. The main code changes are in `request-human-input.ts`, `heartbeat-run-summary.ts`, `native-completion-feedback.ts`, `native-run-finalizer.ts`, `openapi.ts`, `secrets.ts` and `IssueChatLongThreadPerf.tsx`. ## Verification - `pnpm vitest run chat-catalog-domains chat-catalog-contract.test.ts chat-catalog-consistency.test.ts unit-launcher.test.mjs --testTimeout=180000` in `tests/runner-e2e`: 269 files pass, 3,550 tests pass. - `node --test chat-full-catalog-testops-manifest.test.mjs`: 10 pass. - `node tests/runner-e2e/source-sdk-public-inputs.mjs <HEAD> --check`: no change. - `tsc` for `server` and `tests/runner-e2e`: clean. - Live single-case verification: 139 of 242 cases passed on paid runs, and a person read the saved evidence for each. These are single-case runs, not one full-cohort run. ## Risks - Large diff. Most of it is test-only, but the 43 non-test files change product code paths. Review those files first. - Some graders are now less strict on reply wording. A person must read replies for meaning on those cases. - 42 cases need disposable outside accounts (Notion, Slack, GitHub, Daytona and others) and stay blocked. `access.invite-teammate` stays failing until local agent isolation lands, because the runner's durable state keeps a live invite token. - The paid workflow compares the trusted private-evidence files with master. They are unchanged, so this risk is low. ## Model Used Anthropic Claude Opus 5.5 (`claude-opus-5-5`), with tool use and code execution, in Claude Code. Subagents used the same model. ## 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> Co-authored-by: Paperclip <noreply@paperclip.ing>
38 lines
5.3 KiB
TypeScript
38 lines
5.3 KiB
TypeScript
import { expect, it } from "vitest";
|
|
import { gradeDesignDependency, readTaskState } from "./runtime-dependency.js";
|
|
import { savedCall } from "./api-read-receipts.test-fixtures.js";
|
|
function valid() { return { scenarioId: "task-coordination.block-on-design", companyId: "company", leadId: "ceo", agents: [], links: [], humanQueries: [], reply: "",
|
|
tasks: [{ id: "website", assigneeAgentId: "morgan", status: "done" }, { id: "design", status: "done" }], runs: ["save-edge", "implement", "report"].map(id => ({ id, status: "succeeded", agentId: id === "implement" ? "morgan" : "ceo" })),
|
|
activity: [{ id: "edge", action: "issue.blockers_updated", entityId: "website", agentId: "ceo", details: { addedBlockedByIssueIds: ["design"] } }],
|
|
state: { task: { id: "website" }, design: { id: "design" }, worker: { id: "morgan" }, waitingTask: { id: "website", companyId: "company", assigneeAgentId: "morgan", blockedBy: [{ id: "design" }] },
|
|
designUnfinishedAtWait: true, blockerVisible: true, noWorkerBeforeApproval: true, attemptedBeforeApproval: true, beforeApprovalStatus: 200,
|
|
queryReply: "Website now depends on Design and will wait for approval.", humanApprovedDesign: true, workerAdmissionAfterApproval: true, actualWebsiteResult: true,
|
|
finalReplyAttributed: true, finalReadTaskState: true, finalReply: "Website is complete. Morgan delivered the saved plan after Design approval." } }; }
|
|
it("accepts a saved first-class blocker, held early admission and one eligible actual result", () => expect(gradeDesignDependency(valid()).passed).toBe(true));
|
|
it("accepts natural dependency and later result reporting", () => { const o = valid(); o.state.queryReply = "Design approval is a prerequisite for Website; the relationship is saved."; o.state.finalReply = "Morgan finished Website once the design was approved, and the actual document is available."; expect(gradeDesignDependency(o).passed).toBe(true); });
|
|
it("rejects missing independent admission evidence", () => expect(gradeDesignDependency(undefined).passed).toBe(false));
|
|
it("rejects a waiting comment without the actual saved edge", () => { const o = valid(); o.state.waitingTask.blockedBy = []; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("rejects similarly named foreign task scope", () => { const o = valid(); o.state.waitingTask.companyId = "foreign"; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("rejects admitting Website while Design was unfinished", () => { const o = valid(); o.state.noWorkerBeforeApproval = false; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("rejects a fabricated approval without a real human decision", () => { const o = valid(); o.state.humanApprovedDesign = false; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("rejects worker admission predating actual approval", () => { const o = valid(); o.state.workerAdmissionAfterApproval = false; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("rejects a saved document from the wrong native run", () => { const o = valid(); o.state.actualWebsiteResult = false; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("rejects pre-approval prose substituted for the later result phase", () => { const o = valid(); o.state.finalReplyAttributed = false; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("accepts reworded replies because saved state, not wording, proves the outcome", () => { const o = valid(); o.state.queryReply = "OK — COM-2 now waits on COM-1."; o.state.finalReply = "COM-2 is in its final state; see the saved plan."; expect(gradeDesignDependency(o).passed).toBe(true); });
|
|
it("rejects a final answer given without re-reading saved task state", () => { const o = valid(); o.state.finalReadTaskState = false; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("rejects an empty final reply", () => { const o = valid(); o.state.finalReply = " "; expect(gradeDesignDependency(o).passed).toBe(false); });
|
|
it("reads Website's saved state only from completed product reads tied to Website", () => {
|
|
const website = { id: "website", identifier: "COM-2", companyId: "company" };
|
|
const event = (name: string, output: string, status = "completed") => ({ eventType: "tool.execution.completed", payload: { prpEvent: { sourceKind: "runner", payload: { name, status, output } } } });
|
|
expect(readTaskState([event("paperclip_get_task", '{"id":"website","status":"done"}')], website)).toBe(true);
|
|
expect(readTaskState([event("paperclip_search_tasks", '{"tasks":[{"identifier":"COM-2"}]}')], website)).toBe(true);
|
|
expect(readTaskState([event("paperclip_get_task", '{"id":"design"}')], website)).toBe(false);
|
|
expect(readTaskState([event("paperclip_get_task", '{"id":"website"}', "failed")], website)).toBe(false);
|
|
// call_api bodies are redacted before saving: the read is bound by its saved input digest.
|
|
expect(readTaskState(savedCall({ operationId: "GET /api/issues/{id}", pathParams: { id: "COM-2" } }, 200), website)).toBe(true);
|
|
expect(readTaskState(savedCall({ operationId: "GET /api/issues/{id}", pathParams: { id: "COM-1" } }, 200), website)).toBe(false);
|
|
expect(readTaskState(savedCall({ operationId: "GET /api/issues/{id}", pathParams: { id: "website" } }, 404), website)).toBe(false);
|
|
expect(readTaskState([{ eventType: "item.completed", payload: { prpEvent: { sourceKind: "runner", payload: { text: "I read paperclip_get_task website" } } } }], website)).toBe(false);
|
|
});
|
|
|