skills/guillevc/skills/record/Gen Agent Trust Hub

record

Pass

Audited by Gen Agent Trust Hub on Jun 29, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface by design, as its primary function is to process user input and write it to documentation files.
  • Ingestion points: User-supplied facts and decision details (SKILL.md).
  • Boundary markers: Absent; input is formatted into standard markdown structures without additional delimiters.
  • Capability inventory: File-writing access to project documentation paths such as docs/GLOSSARY.md, docs/decisions/, and CLAUDE.md (SKILL.md).
  • Sanitization: No explicit sanitization or escaping of markdown syntax within the user input is defined.
  • Mitigation: The skill mandates a human confirmation step ('Hard gate') before committing Architecture Decision Records (ADRs).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 29, 2026, 02:48 AM
Security Audit — agent-trust-hub — record