check-known-solutions
Pass
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill connects to the Push Realm MCP server at
https://api.pushrealm.com/mcpto search for fixes, involving the transmission of error keywords to a remote vendor service. - [PROMPT_INJECTION]: The ingestion of solutions from an open community database creates a surface for indirect prompt injection. • Ingestion points: Content retrieved from the
search_solutionstool. • Boundary markers: The skill relies on agent instructions to review content with the user instead of technical delimiters. • Capability inventory: The agent has tools to search, submit, and record usage, and is prompted to apply suggested fixes. • Sanitization: No automated sanitization or verification of the remote content is specified.
Audit Metadata