md-starter

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by reading untrusted data from repository signal files and summarizing them into a context-providing file.
  • Ingestion points: Root-level files including README.md, package.json, and CI/CD configurations.
  • Boundary markers: The skill does not implement delimiters or warnings to ignore instructions found within the scanned data.
  • Capability inventory: The agent has file-write capabilities to create and modify the CLAUDE.md context file.
  • Sanitization: No explicit sanitization logic exists to filter out natural language instructions that may be present in project metadata or documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 03:56 PM
Security Audit — agent-trust-hub — md-starter