research-add-items
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's workflow is transparent and limited to its stated research objectives. No evidence of obfuscation, credential theft, or unauthorized network activity was detected.\n- [PROMPT_INJECTION]: The skill includes an indirect prompt injection surface as it processes external content, though this is typical for research tools:\n
- Ingestion points: The workflow reads 'outline.yaml' from the local filesystem and retrieves information from web search results in Step 2.\n
- Boundary markers: Absent; there are no specific markers or instructions to isolate the ingested data from the agent's command logic.\n
- Capability inventory: The skill has the capability to read and write local files and execute web searches via subordinate agents.\n
- Sanitization: Absent; the skill appends external research data directly into the outline file without a validation or cleaning phase.
Audit Metadata