mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-13 15:55:46 +01:00
DottaandClaude
6892ed3a98
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work > - The restart-chaos harness drives the board UI and clicks Retry the way a person would > - When no visible Retry appears, the harness grades `ui_retry_button_missing` and keeps only the banner text > - That is not enough to tell a product bug (no button, a hidden button, a wrong name) from a harness locator bug after the wave has ended > - This pull request saves a full-page screenshot and the banner and Retry-button DOM when the wait fails > - The benefit is that every `ui_retry_button_missing` can be diagnosed from the wave folder ## Linked Issues or Issue Description Refs #20616 **What happened?** Wave `w-20261110T174930-s9205` t01 graded `ui_retry_button_missing`, but the captured banner text contained "Retry". The harness saved no screenshot or DOM, so nobody could prove whether the page had a hidden duplicate or no actionable button. **Expected behavior** A failed Retry-button wait leaves enough evidence to diagnose it later. **Steps to reproduce** 1. Run a restart-chaos wave with the board UI. 2. Get a task whose Retry wait times out. 3. Observe that the wave folder holds only the banner text. ## What Changed - `tests/restart-chaos/board-ui.ts`: add `saveRetryMissingDiagnostics`. On a failed Retry wait it saves `ui-retry-missing-<issue>-<time>.png` (full page) and a `.json` file with the page URL, each `[role=status]`/`[role=alert]` banner's `outerHTML`, and every Retry-named button's visibility, bounding box, banner membership and HTML. Diagnostics never change the click result. - `tests/restart-chaos/main.ts`: the wave's board helper writes to `<wave>/ui-retry-diagnostics`. - `tests/restart-chaos/observer.ts`, `types.ts`: the task's retry attempt records the saved file paths. - `tests/restart-chaos/board-ui.test.ts`: a test for the saved files. ## Verification - `npx vitest run --config tests/restart-chaos/vitest.config.ts tests/restart-chaos/board-ui.test.ts`: 9 passed (new: "saves a screenshot and the banner's HTML when no visible Retry appears"). - `npx tsc --noEmit -p tests/restart-chaos/tsconfig.json` passes. ## Risks - Harness only; no product code changes. A failed screenshot or file write is ignored and does not change the grade. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - Claude Opus 5.5 (`claude-opus-5-5`), Anthropic, used through Claude Code with tool use and code execution. ## 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 <noreply@anthropic.com>