rick-fix

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8). It ingests data from a markdown report file and interpolates fields like {ISSUE} and {FIX_SUGGESTION} directly into the FIX-PROTOCOL.md system prompt and the /tdd skill invocation.
  • Ingestion points: Findings are parsed from the file specified in REPORT_PATH.
  • Boundary markers: No delimiters or 'ignore instructions' warnings are used when interpolating the report content.
  • Capability inventory: The agent can modify source code, edit local markdown reports, and execute shell commands via a test runner.
  • Sanitization: No validation or escaping is performed on the ingested report fields.
  • [COMMAND_EXECUTION]: The skill dynamically executes shell commands to run project tests (Step 4c). It constructs the command using a runner derived from package.json (e.g., npx vitest run) and targets sibling test files found on the local filesystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 04:17 PM
Security Audit — agent-trust-hub — rick-fix