mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-13 15:38:22 +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>
61 lines
5.6 KiB
TypeScript
61 lines
5.6 KiB
TypeScript
import { createHash } from "node:crypto";
|
|
import { mkdtemp, mkdir, writeFile, readFile, realpath, rm, symlink, stat } from "node:fs/promises";
|
|
import path from "node:path";
|
|
import os from "node:os";
|
|
import { afterEach, expect, it } from "vitest";
|
|
import { copyPreparedSourceSdkAudit, retainPreparedSourceSdk } from "./source-sdk-evidence.js";
|
|
import { copySharedEvidence } from "./shared-evidence.js";
|
|
import { inspectCellReportProducer } from "./manual-private-report-assets.js";
|
|
import { stagePrivateDiagnostics } from "./private-diagnostics.js";
|
|
const roots:string[]=[],id="chat-features.runner-opencode.local.plugins.configure-plugin",scenario="plugins.configure-plugin";
|
|
afterEach(async()=>{await Promise.all(roots.splice(0).map(root=>rm(root,{recursive:true,force:true})));});
|
|
const hash=(v:string|Buffer)=>createHash("sha256").update(v).digest("hex");
|
|
async function fixture(){
|
|
const root=await realpath(await mkdtemp(path.join(os.tmpdir(),"sdk-audit-capture-")));roots.push(root);
|
|
const temp=path.join(root,"attempt"),repo=path.join(root,"repo"),handoff=path.join(root,"handoff"),capture=path.join(handoff,"capture"),cell=path.join(capture,"cases",scenario);
|
|
await mkdir(temp,{mode:0o700});await mkdir(handoff,{mode:0o700});await mkdir(cell,{recursive:true,mode:0o700});
|
|
await mkdir(path.join(temp,"selected-private-inputs"),{mode:0o700});
|
|
const secret="synthetic-private-sdk-canary-value";stagePrivateDiagnostics(temp,[id],[secret]);
|
|
for(const directory of ["packages/shared/dist","packages/plugins/sdk/dist"]){
|
|
const dir=path.join(repo,directory);await mkdir(dir,{recursive:true});const body=Buffer.from(secret);
|
|
await writeFile(path.join(dir,"index.js"),body);const outputs=createHash("sha256").update(JSON.stringify(["index.js",body.length])).update(body).digest("hex");
|
|
await writeFile(path.join(dir,".paperclip-build-complete"),JSON.stringify({sources:"a".repeat(64),outputs}));
|
|
}
|
|
const context={root:handoff,private:true as const,runId:"123",attempt:"1",executionIds:[id],cellReportProducer:await inspectCellReportProducer()};
|
|
const source=path.join(temp,"artifacts-private/cases",scenario,"snapshots/source-sdk-privacy-observations.json"),destination=path.join(cell,"snapshots/source-sdk-privacy-observations.json");
|
|
const reject=()=>retainPreparedSourceSdk({repositoryRoot:repo,temporaryRoot:temp,executionIds:[id],sourceRevision:"a".repeat(40),installed:false});
|
|
const copy=()=>copySharedEvidence(capture,cell,temp,[secret],{executionId:id,context});
|
|
return {root,temp,repo,handoff,capture,cell,context,source,destination,secret,reject,copy};
|
|
}
|
|
it("actual SDK privacy rejection keeps its existing typed audit in the external private bootstrap-failure capture",async()=>{
|
|
const f=await fixture();await expect(f.reject()).rejects.toThrow("Source SDK evidence privacy check failed; bytes withheld");
|
|
const bytes=await readFile(f.source);expect(JSON.parse(bytes.toString())).toMatchObject({complete:false,audit:{observations:expect.arrayContaining([expect.objectContaining({classification:"private_exact"})])}});
|
|
await f.copy();expect(await readFile(f.destination)).toEqual(bytes);expect((await stat(f.destination)).mode&0o777).toBe(0o600);
|
|
expect((await readFile(f.destination,"utf8"))).not.toContain(f.secret);await expect(stat(path.join(f.cell,"source-sdk-build"))).rejects.toThrow();
|
|
});
|
|
|
|
it.each(["symlink file","foreign capture","wrong execution","wrong producer","private canary","unknown shape","oversize audit"])("rejects %s before any external audit write",async fault=>{
|
|
const f=await fixture();await expect(f.reject()).rejects.toThrow("Source SDK evidence privacy check failed");
|
|
let cell=f.cell,input={executionId:id,context:f.context};
|
|
if(fault==="symlink file"){const outside=path.join(f.root,"outside.json");await writeFile(outside,await readFile(f.source));await rm(f.source);await symlink(outside,f.source);}
|
|
if(fault==="foreign capture"){cell=path.join(f.root,"foreign");await mkdir(cell);}
|
|
if(fault==="wrong execution")input={...input,context:{...f.context,executionIds:["chat-features.runner-opencode.local.plugins.disable-plugin"]}};
|
|
if(fault==="wrong producer")input={...input,context:{...f.context,cellReportProducer:{...f.context.cellReportProducer,staticManifestSha256:"0".repeat(64)}}};
|
|
if(fault==="private canary"){const audit=JSON.parse(await readFile(f.source,"utf8"));audit.private=f.secret;await writeFile(f.source,JSON.stringify(audit));}
|
|
if(fault==="unknown shape"){const audit=JSON.parse(await readFile(f.source,"utf8"));audit.path="private-unapproved-path";await writeFile(f.source,JSON.stringify(audit));}
|
|
if(fault==="oversize audit")await writeFile(f.source,Buffer.alloc(1024*1024+1));
|
|
await expect(copyPreparedSourceSdkAudit(f.temp,cell,[f.secret],input)).rejects.toThrow();
|
|
await expect(stat(f.destination)).rejects.toThrow();
|
|
});
|
|
it("keeps the captured audit immutable and leaves non-SDK/missing-audit paths unchanged",async()=>{
|
|
const f=await fixture();await f.copy();await expect(stat(f.destination)).rejects.toThrow();
|
|
await expect(copyPreparedSourceSdkAudit(f.temp,f.cell,[],{executionId:"chat-features.runner-opencode.local.tasks.ambiguous-task",context:f.context})).resolves.toBeUndefined();
|
|
await expect(f.reject()).rejects.toThrow("Source SDK evidence privacy check failed");await f.copy();const original=await readFile(f.destination);
|
|
await expect(f.copy()).rejects.toThrow();expect(await readFile(f.destination)).toEqual(original);
|
|
});
|
|
|
|
it("rejects a malformed present framework report instead of hiding it as a missing setup report",async()=>{
|
|
const f=await fixture();await writeFile(path.join(f.capture,"framework-report.json"),"{");
|
|
await expect(f.copy()).rejects.toThrow();
|
|
});
|