break-fix
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The utility script
scripts/quality_gates.pyinvokes shell commands via the Pythonsubprocessmodule to interact with the local Git repository. These operations (e.g.,git diff,git log,git ls-files) are used to perform static analysis on the codebase, identifying unwired exports, unread configuration keys, and coding hygiene issues in the current worktree. - [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves ingesting and analyzing untrusted data from the target application, including source code, network traffic logs, and browser console output. This constitutes a known attack surface where a malicious target could attempt to influence the agent's logic through embedded instructions. The skill instructions mitigate this risk by emphasizing authorized targets and providing explicit rules of engagement.
Audit Metadata