spec-reviewer

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [NO_CODE]: The skill is composed entirely of markdown instructions and configuration. It does not include any scripts, executables, or package dependencies, which significantly reduces the potential attack surface.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for analyzing untrusted external data (product specification files), which is a common vector for indirect prompt injection.
  • Ingestion points: The agent is instructed to "Read the target spec file" in the workflow section of SKILL.md.
  • Boundary markers: None are defined in the instructions to separate the specification content from the agent's internal instructions.
  • Capability inventory: The skill is limited to performing analysis and generating a markdown-formatted scorecard. It has no capabilities for file system modification, network communication, or shell command execution.
  • Sanitization: There are no instructions for sanitizing or escaping the content of the processed files. However, the lack of dangerous tools makes this surface low-risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 06:28 AM
Security Audit — agent-trust-hub — spec-reviewer