test-autofix
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands for test stack detection and execution. This process is protected by a mandatory human-in-the-loop bootstrap phase where users must review and approve all proposed surface commands before they are executed. Additionally, binary resolution checks are performed on every entry to ensure environment consistency.
- [SAFE]: The skill implements 'Hard Refusals' in its rule set, which strictly prevents the agent from using common shortcuts that would weaken the test suite, such as deleting tests, adding skip flags, or removing assertions.
- [SAFE]: A regression-detection mechanism is triggered after every fix; if a change causes new failures, the skill automatically reverts the file to its last committed state using 'git restore'.
- [SAFE]: The workflow uses a formal classification system for failures and a confidence-gate mechanism. High-confidence scores from an independent analysis skill are required before any automated code edits are applied, minimizing the risk of incorrect or speculative fixes.
Audit Metadata