You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
test(developer-experience): port eval cases S7 and N1 #10006
Port scenarios S7 and N1 from the committed sources into plugins/developer-experience/evals/<case>/ (prompt.md, case.yaml, graders/), deterministic graders where possible and llm graders calibrated against labeled samples. Include the rm and compound-shell checks where a case needs them.
Proposed work class: work-class: scoped
Basis: automated intake triage; the lanes App applied this class.
Size: Small
Complexity: Medium
Uncertainty: Knowable
Flags: none
Blocked by: none
Understood: port eval scenarios S7 and N1 into plugins/developer-experience/evals/ as case directories (prompt.md, case.yaml, graders/), using deterministic graders where possible and calibrated llm graders otherwise, including the rm and compound-shell checks. Done when validate-cases.py exits 0, both case.yaml files are tracked, and ci-status is green. The scope is clear. The only open point is locating the committed S7 and N1 sources, referenced via #6649. I found no likely duplicates.
The token broker refused the work-resolve-issue run that agent-ready started because the lanes kill switch was on, so it did no work. work-unblock removed and added agent-ready back to start a fresh run. A kill-switch refusal is no fault in this issue or the lanes, so it never counts against later re-fires; after limits.limit-stops-per-issue of them since a person last added agent-ready or a run last reported, work-unblock waits longer before each one, up to 24 hours.
The work-resolve-issue lane did not start this issue: it needs a person to make it clearer first.
Why, in the run's own words:
The scenario definitions for S7 and N1 are not in the repository, so there is nothing to port and no way to tell what either case should test. The issue says to port them 'from the committed sources', but plugins/developer-experience/evals/ does not exist on main (no sources/ directory, no S7 or N1 text anywhere in the checkout). Parent #6649 says the scenarios live only in the gitignored .work/devx/evals/ harness on the maintainer's host (PR #6668's worktree). Committing them to plugins/developer-experience/evals/sources/ is #9936, which is open and labeled needs-human. The sibling issue #9979 (the same S7/N1 port) lists #9936 and #9943 (shared fixtures) as blockers, and this issue does not. Before this can run: a person lands #9936 (and #9943 if the cases need the shared fixtures), or pastes the S7 and N1 scenario definitions, checklists and labeled samples into this issue. Also confirm whether #10006 duplicates #9979 and which one should stay open.
What the run says it did not do:
Port S7 and N1 into plugins/developer-experience/evals/<case>/: blocked because their scenario sources are not committed (#9936 open)
python3 plugins/evals/skills/validate/scripts/validate-cases.py plugins/developer-experience/evals: not run, because no cases were written and this lane cannot run an interpreter
ci-status green on the PR: left to the PR's CI, and no PR results from a pushback
Activity
github-actions commentedon Oct 25, 2026
Proposed work class: work-class: scoped
Basis: automated intake triage; the lanes App applied this class.
Size: Small
Complexity: Medium
Uncertainty: Knowable
Flags: none
Blocked by: none
Understood: port eval scenarios S7 and N1 into plugins/developer-experience/evals/ as case directories (prompt.md, case.yaml, graders/), using deterministic graders where possible and calibrated llm graders otherwise, including the rm and compound-shell checks. Done when validate-cases.py exits 0, both case.yaml files are tracked, and ci-status is green. The scope is clear. The only open point is locating the committed S7 and N1 sources, referenced via #6649. I found no likely duplicates.
Automated triage, applied by the lanes App.
melodic-automation-lanes commentedon Oct 25, 2026
The token broker refused the work-resolve-issue run that
agent-readystarted because the lanes kill switch was on, so it did no work. work-unblock removed and addedagent-readyback to start a fresh run. A kill-switch refusal is no fault in this issue or the lanes, so it never counts against later re-fires; after limits.limit-stops-per-issue of them since a person last addedagent-readyor a run last reported, work-unblock waits longer before each one, up to 24 hours.melodic-automation-lanes commentedon Oct 25, 2026
The work-resolve-issue lane did not start this issue: it needs a person to make it clearer first.
Why, in the run's own words:
What the run says it did not do:
Run: https://github.com/melodic-software/claude-code-plugins/actions/runs/40223407414
To retry, answer the question here, remove
needs-humanand addagent-readyagain.kyle-sexton commentedon Oct 25, 2026
Duplicate split of #6649 by a second Backlog HITL writer. To reopen: reopen this issue and say what the kept child misses.
Kept child: #9979.