quality-gates-parallel
Pass
Audited by Gen Agent Trust Hub on May 3, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection (Category 8).
- Ingestion points: Reads and processes untrusted source code files (e.g., 'src/auth.ts') and aggregates JSON results from subagents located in the '.claude/quality-results/' directory.
- Boundary markers: Absent. The skill instructions do not specify the use of delimiters or instructions to ignore embedded content within the files being analyzed.
- Capability inventory: The skill has access to high-privilege tools including 'Bash', 'Write', and the native Claude 'Task' management tools ('TaskCreate', 'TaskUpdate', 'TaskGet').
- Sanitization: There is no documented logic for sanitizing or validating the content of the target files or the output of subagents before they are used to make automated decisions, such as blocking merges.
- [COMMAND_EXECUTION]: The skill utilizes the 'Bash' tool to execute local orchestration scripts, specifically '.claude/scripts/quality-coordinator.sh' and '.claude/scripts/read-quality-results.sh', as well as specialized subagent commands like '/sec-context-depth' and '/code-review'. This is consistent with its purpose as an automation orchestrator.
Audit Metadata