skills/kensaurus/skills/docs-adr/Gen Agent Trust Hub

docs-adr

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is focused on repository documentation and maintaining a record of architectural choices. It does not include hardcoded credentials, network operations, or privileged command execution.
  • [INDIRECT_PROMPT_INJECTION]: The skill establishes a process where the agent reads and adheres to architecture decisions stored in markdown files. Ingestion points: The agent reads the docs/adr/ directory and INDEX.md to inform its proposals (SKILL.md, Phase 0). Boundary markers: The skill uses a structured markdown format but does not implement explicit boundary markers or instruction isolation for the content of the ADRs. Capability inventory: The agent is authorized to read and write markdown files within the repository's documentation path. Sanitization: There is no specified sanitization or validation for the content of the ADRs. However, the skill includes a human-in-the-loop requirement where the agent must flag contradictions to the user rather than silently following them (SKILL.md, Phase 4).
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 11:26 PM