protocolsio-integration
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- PROMPT_INJECTION (LOW): The skill allows the agent to ingest and act upon untrusted data from a collaborative platform. Ingestion points: Protocol descriptions and step contents in references/protocols_api.md, and comment bodies in references/discussions.md. Boundary markers: No markers or delimiters are specified to isolate untrusted content from instructions. Capability inventory: The skill enables authenticated network operations for creating, updating, and deleting protocols, comments, and workspace memberships. Sanitization: No data sanitization or input validation is described in the provided documentation.
- CREDENTIALS_UNSAFE (SAFE): The authentication documentation uses safe placeholders (e.g., YOUR_TOKEN) for API secrets and follows standard OAuth2 protocols.
Audit Metadata