Skip to content

fix(tests): mutation_testing_tests MUT-05/06/07 read ci.yml from the crate dir — dark on every box (#3126) - #3128

Closed
noahgift wants to merge 1 commit into
mainfrom
PMAT-1098-3126-mut-paths
Closed

fix(tests): mutation_testing_tests MUT-05/06/07 read ci.yml from the crate dir — dark on every box (#3126)#3128
noahgift wants to merge 1 commit into
mainfrom
PMAT-1098-3126-mut-paths

Conversation

@noahgift

Copy link
Copy Markdown
Contributor

Closes #3126.

crates/aprender-core/tests/includes/mut0.rs resolved .github/workflows/ci.yml with a repo-relative Path::new(...), but cargo and nextest run every test binary with cwd = the crate dir (crates/aprender-core), so MUT-05/06/07 could not pass on any box — the target was dark until the quick tier ran it (measured 2026-09-11 15:47Z, cargo nextest run -p aprender-core --tests: 3 failed). Every repo path in the file now anchors on CARGO_MANIFEST_DIR/../.. (the monorepo_invariants.rs pattern; same class as the #3112 fixes for falsification_model_oracle_tests, realizar_integration_tests, toyota_principles_tests), and each panic prints the path it tried.

Written by an agy goal lane (PMAT-1098 phase 3, conversation in the receipt); re-verified by the orchestrator:

cargo nextest run -p aprender-core --test mutation_testing_tests --no-fail-fast
     Summary [   0.025s] 16 tests run: 16 passed, 0 skipped
bash scripts/check_tree_reader_tests.sh   -> exit 0 (registry unchanged)
cargo fmt -p aprender-core -- --check     -> exit 0

No roadmap change on purpose: this PR stays mergeable after the 0.67 train (#3127) squashes.

🤖 Generated with Claude Code

…crate dir — dark on every box (#3126)

Tests mut05_ci_mutation_workflow_exists, mut06_mutation_artifacts_captured, and mut07_mutation_timeout_configured in mutation_testing_tests were failing because they used repo-relative paths (`.github/workflows/ci.yml`) while cargo nextest runs them from `crates/aprender-core`. This anchors the paths to `CARGO_MANIFEST_DIR` resolving from the workspace root.

Pmat-Ticket: PMAT-1098

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
@noahgift noahgift added this to the 0.67.0 milestone Sep 11, 2026
@noahgift noahgift added bug Something isn't working pp-067 PP-066 spec §6: 0.67 lane (carried row) labels Sep 11, 2026
@noahgift
noahgift enabled auto-merge September 11, 2026 19:31
@noahgift
noahgift disabled auto-merge September 11, 2026 19:35
@noahgift

Copy link
Copy Markdown
Contributor Author

Auto-merge disarmed until the 0.67 train (#3127, carries the #3112 aarch64 lint fix) lands: gx10 pools carry clean-room now and this branch is off main @ bbd3243 without that fix, so a merge-group ci / lint on gx10 would evict it. Re-arm after gh pr update-branch 3128.

@github-actions

Copy link
Copy Markdown

§13.11 rung 1 — quorum shadow verdict

S13-SHADOW pr=3128 head=04a1c80c74cf86f4d5fee11efbbaaa763005ca5a verdict=REFUSE class=Q1 arm_rc=1

Shadow mode: this records a verdict and merges nothing. A refusal
to arm is not a block (§13 adds zero rows to §7) — the pull request is
exactly as green as it was.

@noahgift

Copy link
Copy Markdown
Contributor Author

Superseded: main already carries this fix — train #3127 (cb829fc) landed mut_workspace_root() for MUT-05/06/07 in crates/aprender-core/tests/includes/mut0.rs via the #3131 sweep; this branch's remaining diff is the same intent expressed with an explicit canonicalize. Closing #3126 with the sha.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

bug Something isn't working pp-067 PP-066 spec §6: 0.67 lane (carried row)

Projects

None yet

1 participant