security-review

Pass

Audited by Gen Agent Trust Hub on Jul 26, 2026

Risk Level: SAFE
Full Analysis
  • [DYNAMIC_CONTEXT_INJECTION]: The skill uses !date and !git rev-parse --short HEAD to include metadata at load time. These are benign commands typical for development workflows.
  • [INDIRECT_PROMPT_INJECTION]: The skill performs full-repository security reviews, creating a surface for indirect prompt injection where malicious instructions in the codebase could influence the agent's output.
  • Ingestion points: Full repository content (source code, configuration files, CI/CD pipelines) as described in the review workflow.
  • Boundary markers: Absent; the skill does not use delimiters to isolate untrusted repository data from its instructions.
  • Capability inventory: The skill utilizes the obsidian_append_content tool to write reports to a local Obsidian vault.
  • Sanitization: No evidence of sanitization or escaping of the ingested code before it is summarized in the report.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 26, 2026, 12:10 AM
Security Audit — agent-trust-hub — security-review