tracer-implement

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from 'requests, tickets, or specs' to define implementation tasks, creating an attack surface where malicious input could influence agent behavior.\n
  • Ingestion points: The goal and acceptance requirements are derived from external sources in SKILL.md (Step 2) and recorded in plan.md.\n
  • Boundary markers: There are no explicit delimiters or 'ignore embedded instructions' warnings for external content in the implementer-prompt.md or fixer-prompt.md templates.\n
  • Capability inventory: The agent has capabilities to write files, execute Git operations via scripts (scripts/review-package, scripts/workspace), and run shell commands for testing as defined in SKILL.md.\n
  • Sanitization: No evidence of input validation or content sanitization is present for the ingested ticket data.\n- [DYNAMIC_EXECUTION]: The implementation workflow (Step 3 in SKILL.md) requires the agent to 'self-check changed behavior' by running 'focused red/green tests'. This involves the dynamic execution of code or scripts generated by the agent during the session to verify its changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 11:23 PM
Security Audit — agent-trust-hub — tracer-implement