anchor-authoring
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill involves processing external repository documentation and user communications to generate authoring outcomes.
- Ingestion points: The agent is instructed in SKILL.md to 'Read applicable existing product intent, requirements, design, and decisions through the repository's knowledge system.'
- Boundary markers: No specific delimiters or boundary markers for ingested text are defined in the instructions.
- Capability inventory: The skill enables reading existing repository files and authoring new Anchor artifacts.
- Sanitization: While no technical sanitization is described, the 'QA.md' file provides a comprehensive adversarial check for 'false authority' and 'hidden contradictions' to validate the generated output against source intent.
Audit Metadata