review-spec

Pass

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill handles untrusted external content ($ARGUMENTS) without using delimiters or boundary markers, creating a surface for indirect prompt injection.
  • Ingestion points: Technical documents are passed to the tech-spec-reviewer subagent through the $ARGUMENTS variable.
  • Boundary markers: Absent; the prompt in SKILL.md does not use delimiters to isolate the document content from instructions.
  • Capability inventory: The skill can read files and execute git/node bash commands, providing a path for potential abuse if an injected instruction is followed by the agent.
  • Sanitization: No input validation or sanitization is implemented for the processed documents.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 20, 2026, 07:24 PM
Security Audit — agent-trust-hub — review-spec