Files
PaperClipAI/tests/runner-e2e/chat-catalog-domains/runtime-managed-profile.test.ts
T
f4337bb5fb test(evals): land the chat catalog eval harness on master (#19614)
## 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>
2026-11-10 08:02:11 -06:00

31 lines
5.2 KiB
TypeScript

import { expect, it } from "vitest";
import { gradeManagedProfile, runtimeManagedProfileCases } from "./runtime-managed-profile.js";
function valid() {
const agent = { id: "ceo", companyId: "company", adapterType: "paperclip_runner", adapterConfig: { provider: "codex" } };
return { scenarioId: "agents.managed-profile", companyId: "company", leadId: "ceo", agents: [agent], tasks: [], activity: [], links: [], humanQueries: [],
reply: "Only a board user can do this: run `paperclipai managed-agent setup --profile-key <key> --display-name <name> --api-key-secret-id <id> --acknowledge-retention`.",
runs: [{ id: "guidance", status: "succeeded" }], state: { profilesBefore: [], profilesAfter: [] as unknown[], agentsBefore: [structuredClone(agent)],
agentReadStatus: 403, agentProvisionStatus: 403, operatorInventoryStatus: 200, operatorContractAvailable: true,
retentionProbeStatus: 422, retentionProbeError: "Enabling Managed Agents requires the retention acknowledgement", qualificationProbeStatus: 422,
qualificationProbeError: "Claude Managed qualification attestation is required before enablement", operatorResourceVisible: true } };
}
it("accepts the actual no-profile context with a reference to where the board user acts", () => expect(gradeManagedProfile(valid()).passed).toBe(true));
// Recorded reply from run 42596676991: it names the board-only profile route it was denied on.
const RECORDED_2 = "I can't create the Anthropic managed agent from this chat. Claude Managed Agent profiles are board-only in Paperclip: I verified that `GET /api/companies/{companyId}/managed-agent-profiles` returns **403 \"Board access required\"**, which is final for agent credentials. The company currently has only the CEO agent.\n\n**Blocker owner:** Board (you). **Unblock action:** a board user must create and attest the profile in the Paperclip web app:\n1. Store an Anthropic API key as a company secret on [Secrets](/COM/company/settings/secrets).\n2. Create the agent in Anthropic's managed-agents beta (`managed-agents-2026-04-01`) and note its agent ID/version plus a limited environment (no hosts, no packages).\n3. Create the Claude Managed Agent profile with that secret and environment, acknowledge stateful retention, and enable it. Start from [Anthropic setup](/COM/apps/connect?source=anthropic).\n\nThis setup cannot be completed until that board-only step is done. If you want a regular Claude Code agent instead, I can hire one once an Anthropic connection is authenticated.";
// Reply from run 42594125035: it sent the board to a settings control that does not exist and names neither literal.
const RECORDED_1 = "I can\u2019t create it from this chat: Claude Managed Agent profiles are restricted to board/operator credentials, and agent credentials receive 403.\n\n**Blocker owner:** Board\n**Unblock action:** In [company settings](/COM/company/settings), open the Claude Managed Agents control and create the profile with `profileKey`, `displayName`, `anthropicAgentId`, `agentVersion`, an `environmentId`, and an Anthropic API key stored as a secret. Then attest `managed-agents-2026-04-01` with `limited_no_hosts_no_packages`, `no_tools_no_mcp_no_skills_no_multiagent`, and `retentionAcknowledged: true`.\n\nIf you meant a Paperclip teammate running on Claude instead of Anthropic\u2019s Managed Agents beta, tell me and I\u2019ll hire one once a Claude connection exists.";
it("grades the place to act by product literal, not by wording", () => {
for (const reply of [RECORDED_2, "A board user creates it with the CLI: paperclipai managed-agent setup.", "This needs board credentials on POST /api/companies/{companyId}/managed-agent-profiles."])
{ const o = valid(); o.reply = reply; expect(gradeManagedProfile(o).passed, reply).toBe(true); }
for (const reply of [RECORDED_1, "Ask an operator to set it up in the web app.", ""]) { const o = valid(); o.reply = reply; expect(gradeManagedProfile(o).passed, reply).toBe(false); }
});
it("rejects missing independent management evidence", () => expect(gradeManagedProfile(undefined).passed).toBe(false));
it("rejects quietly provisioning a profile despite ordinary chat authority", () => { const o = valid(); o.state.profilesAfter.push({ id: "unexpected" }); expect(gradeManagedProfile(o).passed).toBe(false); });
it("rejects an ordinary agent being allowed to read or provision", () => {
for (const key of ["agentReadStatus", "agentProvisionStatus"] as const) { const o = valid(); o.state[key] = 200; expect(gradeManagedProfile(o).passed).toBe(false); }
});
it("rejects substituting another runtime configuration", () => { const o = valid(); o.agents[0]!.adapterConfig.provider = "acpx"; expect(gradeManagedProfile(o).passed).toBe(false); });
it("rejects a fabricated supported management path", () => { const o = valid(); o.state.operatorContractAvailable = false; expect(gradeManagedProfile(o).passed).toBe(false); });
it("rejects claiming required qualification without an actual validation guard", () => { const o = valid(); o.state.qualificationProbeStatus = 201; expect(gradeManagedProfile(o).passed).toBe(false); });
it("versions the regraded oracle", () => expect(runtimeManagedProfileCases[0]!.oracleVersion).toBe("paperclip.chat-catalog.runtime.managed-profile.v5"));