tech-hub-mcp
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and metadata are consistent with its stated purpose of managing MCP server lifecycles and optimizing data fetches.- [SAFE]: No credential harvesting, hardcoded secrets, or data exfiltration patterns were identified in the files.- [PROMPT_INJECTION]: Static analysis identified an indirect prompt injection surface (Category 8) due to the skill's role in fetching external data, but no malicious patterns are present. Ingestion points: AGENTS.md (identifying server configs and fetching data from APIs/databases). Boundary markers: Absent. Capability inventory: SKILL.md (activating/deactivating servers), AGENTS.md (using Grep/Glob for project exploration). Sanitization: Not explicitly defined. The behavior is inherent to the skill's utility and does not constitute a finding.
Audit Metadata