qa-gate

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Executes local project scripts (pnpm test, pnpm check-types, pnpm lint) to validate code quality. This is standard behavior for a quality gate tool and uses the well-known pnpm package manager.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by processing external data such as story files and code changes.
  • Ingestion points: Reads documentation from docs/stories/ (Phase 1) and analyzes repository source code during specialist reviews (Phase 3).
  • Boundary markers: Input content is interpolated into sub-agent prompts (haiku-based reviews) without explicit isolation or instructions to ignore embedded commands.
  • Capability inventory: The skill has permissions to write files to the repository (docs/qa/gates/), modify existing story files, and execute subprocesses via pnpm.
  • Sanitization: There is no evidence of sanitization, escaping, or filtering of ingested text before it is presented to the AI sub-agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 09:46 PM