grimoire-scribe

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues detected. The skill performs intended local documentation tasks, including reading repository structure and architecture records to generate specification files. The workflow includes a mandatory human-in-the-loop verification step before any changes are committed to the filesystem, which mitigates risks associated with processing repository content.- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests local configuration files, which is necessary for its primary synthesis purpose.
  • Ingestion points: Reads context and architecture decision records from .grimoire/CONTEXT.md and .grimoire/adr/*.md (SKILL.md).
  • Boundary markers: Absent; the skill relies on the 'Present draft' step for user review before file creation.
  • Capability inventory: Restricted to reading the repository structure and writing markdown files to the .grimoire/spec/ directory.
  • Sanitization: None; the skill synthesizes provided context into a structured document.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 09:10 AM