llm-wiki-builder
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a standard and legitimate workflow for automating research documentation and wiki building.- [PROMPT_INJECTION]: The skill possesses a data ingestion surface area as it processes content from external research databases and local documents.
- Ingestion points: Content retrieved from PubMed Search MCP tools, Zotero library items, and local files (Markdown, PDF, DOCX).
- Boundary markers: The instructions do not explicitly define delimiters to separate ingested external content from the agent's internal instructions.
- Capability inventory: Filesystem access for inspecting project markers and writing Markdown notes; network operations facilitated via Zotero and PubMed tools.
- Sanitization: No explicit sanitization or filtering of external content is described in the skill logic.- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill interacts with research data and local project files strictly according to its primary purpose. No hardcoded credentials, sensitive system path access, or unauthorized network exfiltration patterns were detected.
Audit Metadata