review-prd-dev-gaps
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted content including Product Requirements Documents (PRDs), development plans, milestone records, and Architectural Decision Records (ADRs). This creates an attack surface where malicious instructions embedded in these documents could attempt to influence the agent's report generation or the logic of subsequent plans generated by the
dev-planskill. - Ingestion points: Files located in
docs/,goal/, and specific paths provided as arguments (PRD, plans, records). - Boundary markers: The instructions do not define specific delimiters or "ignore instructions" wrappers for the ingested content.
- Capability inventory: The skill can write files (reports and gaps plans), call other skills (specifically
dev-plan), and interact with the user to triage discrepancies. - Sanitization: No explicit sanitization or filtering of external document content is described in the provided instructions.
- [COMMAND_EXECUTION]: The skill instructions require the agent to investigate the current environment and repository state (e.g., checking git commit hashes, "dirty" status, and environment configurations). This involves executing shell commands to gather metadata, which is consistent with the skill's stated purpose of auditing delivery against a specific code baseline.
Audit Metadata