Files
PaperClipAI/tests
DottaandPaperclip 9e61fdebf9 test(evals): pin approval-link fix for case 564 (#18909)
## Thinking Path

> - Paperclip manages work through agents and explicit approval gates.
> - The hiring revision case requires a usable link to the saved pending
proposal.
> - A rendered unprefixed approval link currently misses the company
redirect.
> - The reviewed fix adds one existing redirect route. It changes no
approval authority.
> - The TestOps workflow reads its candidate source from a protected
master pin.
> - This PR pins that one-fix candidate so one remote case can test the
original journey.

## Linked Issues or Issue Description

Refs #17858. Refs #18780.

The source candidate applies only the approval-detail redirect from the
existing fix. The core fix PR remains unmerged. This PR changes the
evaluation source pin.

## What Changed

- Pin candidate `abbde061c7ec9181021858f7b1af21a48637f97f`, a direct
child of the prior `fd5e3b1e` candidate.
- Update the matching launch snapshot target and the expected-pin test.
- Keep all 244 case rows, original inputs, clocks, model, and private
evidence rules unchanged.

## Verification

- The candidate differs from its parent in one line in `ui/src/App.tsx`.
- The owning browser regression follows a real rendered link to a saved
pending proposal. It fails before the route fix and passes after it. The
18 App tests, UI build, typecheck, and token gates passed.
- The candidate token gates passed. The existing source SDK input check
reports 506 unchanged inputs and unchanged runtime pins. No SDK registry
or certificate was renewed.
- The three existing TestOps contract, adapter, and manifest test files
passed: 47 tests, zero failures.
- The first local pin invocation lacked the YAML dependency. It is
retained as a setup failure. The invocation with the existing locked
dependencies passed without assertion changes.
- Exact-head normal PR CI
[41763017485](https://github.com/paperclipai/paperclip/actions/runs/41763017485)
completed successfully: 48 successful jobs and one selector skip. Fresh
Greptile and Paperclip reviews both report 5/5 with no findings; all
review threads are resolved. The optional Superagent security check is
neutral, so no security assessment PASS is claimed. No new provider run
or TestOps PASS is claimed. The intended next check is only
`questions-and-approvals.formal-hire-revision` on existing plan 102 and
job 100.

## Risks

The route regression does not prove the full agent journey passes. The
first remote run will test that remaining gap.

The candidate keeps the prior catalog application base plus this one
route fix. It does not refresh the whole runtime or qualify all 244
cases. The existing broader qualification holds remain. Automatic model
retries stay zero and the authorized concurrency cap stays eight.

## Model Used

OpenAI `gpt-6.1-sol` through Codex. Capabilities used: source analysis,
code integration, and local validation. The host does not expose the
context-window limit. The model under evaluation remains
OpenCode/OpenRouter `deepseek/deepseek-v4.1-flash-1017`.

## 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
- [x] All Paperclip CI gates are green
- [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-11-03 21:00:36 -06:00
..