Repository navigation
fix(ci): clear the 2 red checks on main at source - #494
hyperpolymath wants to merge 3 commits into
Conversation
`panic-attack assail` is red on main. The scanner at f2a6a70 reports the 2 deliberate `sorry` detector fixtures under tests/corpus_fixtures/ as Critical ProofDrift. Both fixtures already carry an inline `panic-attack: accepted` marker, but f2a6a70 does not anchor a Lean or Isabelle `sorry` finding to its line, so it cannot apply the marker. 55ffcc9 is panic-attack main and includes panic-attack#278, which anchors those findings. bridge-gate.yml reads the same pin. The Patch Bridge source and its report schema (0.3.0) did not change between the 2 commits. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0136P7ZmAsYpFA9gxfQLF3F6 Signed-off-by: Claude <noreply@anthropic.com>
The handler checked `session.state.is_none()` and then called `.unwrap()` on the same Option 4 times. Hypatia reports that as code_safety/unwrap_without_check. The baseline entry for that rule names src/rust/server.rs, which no longer exists, so it does not cover this file. Bind the state with let-else. Read the goal counts from values the handler already holds. A failed tactic leaves the state as it was, so its count is the count from before the call. No behaviour changes. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0136P7ZmAsYpFA9gxfQLF3F6 Signed-off-by: Claude <noreply@anthropic.com>
provers::test_fixture is compiled only under cfg(all(test, unix)), but that gate is on the `mod` line in provers/mod.rs. Hypatia scans the file alone and reports its expect() and panic! calls as code_safety/expect_in_hot_path and code_safety/panic_macro. write_executable now returns io::Result and names the failed step in the error. The 4 callers are inside `#[cfg(test)] mod tests` and call .unwrap(), so a failed fixture still stops the test with the same text. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0136P7ZmAsYpFA9gxfQLF3F6 Signed-off-by: Claude <noreply@anthropic.com>
|
Warning Review limit reachedYou've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Next included review available in 13 minutes. View limit details
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. |
|
Codex Review Summary: ❌ FailedThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
This security review couldn’t start because the paying account has exhausted its Codex allowance and has insufficient available credits. The payer can check their balance and add credits in the Codex usage dashboard, then try again. If you do not manage the paying account, contact this repository's admins. |



Summary
2 checks are red on
mainatb6ff166, and both are older than the recent PRs:Static Analysis / panic-attack assailandgovernance / Validate Hypatia Baseline. This PR fixes each finding at its source. No baseline or ledger entry is added.Changes
static-analysis-gate.yml:PANIC_ATTACK_REFmoves fromf2a6a70to55ffcc9a240e937edd968487af15ab45c6f0522f(panic-attackmain).sorrydetector fixturestests/corpus_fixtures/isabelle/Smoke.thyandtests/corpus_fixtures/minif2f/algebra.lean. Both already carry apanic-attack: acceptedmarker.f2a6a70does not anchor a Lean or Isabellesorryfinding to its line, so it cannot apply the marker. feat(assail): line-anchor Lean/Isabelle sorry findings so inline markers apply panic-attack#278 adds that, and55ffcc9includes it.bridge-gate.ymlreads the same pin.src/bridge/and the Patch Bridge schema (0.3.0) did not change between the 2 commits.src/rust/server/session.rs: the apply-tactic handler checkedstate.is_none()and then called.unwrap()4 times. It now binds the state with let-else and reads the goal counts from values it already holds. No behaviour changes.src/rust/provers/test_fixture.rs: the helper is compiled only undercfg(all(test, unix)), but the gate is on themodline, so Hypatia scans the file alone.write_executablenow returnsio::Resultwith the failed step in the error. The 4 callers in#[cfg(test)] mod testscall.unwrap(), so a failed fixture still stops the test.RSR Quality Checklist
Required
cargo test --lib -- provers::chuffed provers::faial provers::gpuverify provers::version_probe server::gives 58 passed, 0 failed.cargo fmt -- --checkon the 6 Rust files is clean.unsafeblocks without// SAFETY:comments.envfiles includedAs Applicable
Testing
cargo install --git … --rev 55ffcc9… --locked --features httpbuilt locally. On a clean copy ofmain(b6ff166),static-analysis-evidence.sh scan panic-attackgavestatus: ran, exit 0, coverage 951/951. Both fixture findings aresuppressed: true.static-analysis-evidence.sh gateexit 0.bash scripts/ci/test-static-analysis-evidence.shgives 59 passed.lib/rules/code_safety.exat the pinned82396d8(same comment and#[cfg(test)]stripping) to both files. Before:session.rs4unwrap_without_check,test_fixture.rs3expect_in_hot_pathand 2panic_macro. After: 0 for every code_safety Rust rule. The hostedValidate Hypatia Baselinerun on this PR is the real proof.Screenshots
Not applicable.
Generated by Claude Code