ai-implement-spec
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon external specifications and ticket data, which could contain malicious instructions meant to influence agent behavior.
- Ingestion points: The skill reads external data from a 'spec' and 'issue tracker' (tickets) to build a task graph and drive implementation work in Step 1.
- Boundary markers: There are no specific instructions to treat the spec or ticket content as untrusted data or to use delimiters to separate instructions from the data content.
- Capability inventory: The skill orchestrates multiple subagents, creates git branches/worktrees (Step 4), executes code through the 'tdd' skill tool (Step 4), and performs merges (Step 5).
- Sanitization: The instructions do not define any sanitization, validation, or filtering for the data read from the external specifications or tickets.
Audit Metadata