knowledge-base-enrichment

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill follows security best practices by implementing a human-in-the-loop workflow. All proposed changes are staged in a dedicated directory for inspection before they are applied to the knowledge base.- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied Markdown notes, which serves as an ingestion point for potentially untrusted data. * Ingestion points: Reads Markdown source files from the vault/raw/ directory (SKILL.md). * Boundary markers: Explicitly separates source notes (read-only) from synthesized wikis and enrichment proposals. It instructs the agent to read frontmatter without rewriting the main content body. * Capability inventory: Uses file system read and write capabilities within the defined project scope. * Sanitization: Employs a mandatory verification phase and log auditing to detect unresolved links or stale claims before finalization.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 03:53 PM
Security Audit — agent-trust-hub — knowledge-base-enrichment