workflow-fix: add eval_results/issue_1481 to tests/sparse_cones.txt (issue-1947 tests collect in sparse worktrees)
Highlight text on any card (body, plan, or an event) to anchor a comment, or leave a whole-task comment here. Mention @claude to summon a reply.
No comments yet.
Overview / Motivation
Auto-filed by the workflow-fix-on-bug protocol from a prose follow-up raised on task #2021 (emitting agent: implementer, unit A).
Goal
Add eval_results/issue_1481 to tests/sparse_cones.txt so tests/test_issue1947_resume_matrix.py (and the other issue-1947 tests importing scripts/issue1947_cells.py) collect/run in fresh sparse worktrees without a manual git sparse-checkout add.
Workflow gap
- Bug observed:
tests/test_issue1947_resume_matrix.pyfails at collection in any sparse worktree lacking theeval_results/issue_1481cone; the #2021 unit-A implementer had to rungit -C "$WT" sparse-checkout add eval_results/issue_1481by hand (16/16 passed after). - Why it is a workflow gap: CLAUDE.md § sparse-worktree test gate mandates that any test hard-reading
repo_root()/eval_results/issue_<M>/...gets its dir added to thetests/sparse_cones.txtregistry consumed byscripts/new_worktree.sh; a missing line makes every fresh sparse worktree's Step 9c gate red on an untouched test (the #671 class). - Confidence (emitter): high
- verified-at-filing:
grep -n 'issue_1481' tests/sparse_cones.txt→ 0 hits (absence claim — the registry line is missing; 2026-08-02). Call-hop trace (clause g): symptom site =tests/test_issue1947_resume_matrix.py(importsissue1947_datagen→issue1947_cells); construction site =scripts/issue1947_cells.py:43(VERDICT_MANIFEST_PATH = REPO_ROOT / "eval_results/issue_1481/analysis/verdict_manifest.json", read at L113 with a raise naming the manual sparse-checkout remedy at L111). Landed-fix history:git log --oneline --since='7 days ago' -- tests/sparse_cones.txt→ 1 commit (ded60e3fa8, task #1776 — adds a DIFFERENT cone, the issue1776 phase3 manifest; not this fix).
Proposed change (candidate diff sketch — refine in planning)
+ eval_results/issue_1481
(one registry line in tests/sparse_cones.txt; planner should confirm whether sibling issue-1947 test files or eval_results/issue_1947 need lines too, per the registry's own conventions)
Scope / surfaces
- Primary target:
tests/sparse_cones.txt - Grep the workflow surface for the pattern before editing (
grep -rln 'issue_1481' tests/sparse_cones.txt scripts/new_worktree.sh) and update every hit; list them in the plan.
Constraints / invariants
- Workflow-surface only — never experiment code,
configs/, ortasks/. scripts/workflow_lint.py --check-askspasses; ruff on touched files passes; ifworkflow.yamlorCLAUDE.mdchange, they stay consistent with the rule file.- This session runs under
EPM_WORKFLOW_FIX_SESSION=1and carries aworkflow_fix_target:Provenance line — it MUST NOT auto-route any of its own subagents' workflow-fix candidates (recursion guard).
Provenance
- workflow_fix_target: tests/sparse_cones.txt
- fingerprint: 31e37654e526
Surfaced prose (verbatim, from the #2021 unit-A implementer manifest): "tests/test_issue1947_resume_matrix.py fails at COLLECTION in any sparse worktree lacking the eval_results/issue_1481 cone, and eval_results/issue_1481 is absent from tests/sparse_cones.txt — a pre-existing gap unrelated to this diff (I ran git -C "$WT" sparse-checkout add eval_results/issue_1481 locally; 16/16 passed). Unit B / the gate may hit this; the registry line is the documented fix."