wiki-context-pack
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill logic is focused on legitimate data processing of a local Obsidian vault. It resolves configuration from standard locations like .env and ~/.obsidian-wiki/config and restricts its operations to file reading and logging within the vault path.
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by processing potentially untrusted wiki content for consumption by other agents. Ingestion points: Wiki page bodies, index.md, and hot.md in the vault. Boundary markers: Employs markdown headers and horizontal rules as structural delimiters, though it lacks explicit instructions to ignore embedded content. Capability inventory: Accesses local files for reading and appends logs to log.md. Sanitization: Performs content compression and distillation but does not filter for instructions within the processed text.
Audit Metadata