find-people-having-simple-signals

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Interacts with the OpenFunnel API at relative endpoints (e.g., /api/v1/signal/get-signal-list) to retrieve lead and signal data. This behavior is consistent with the skill's stated purpose of tracking professional social listening signals.
  • [COMMAND_EXECUTION]: Guides the agent to manage authentication tokens by writing them to a local .env file and updating .gitignore. This follows industry best practices for local secret management to ensure credentials are not accidentally shared.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it processes data retrieved from external LinkedIn profiles and posts.
  • Ingestion points: External content from LinkedIn profiles ingested via API responses in SKILL.md.
  • Boundary markers: Absent.
  • Capability inventory: Performs network requests to OpenFunnel APIs and writes to the local filesystem (.env) as described in SKILL.md.
  • Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 06:03 AM
Security Audit — agent-trust-hub — find-people-having-simple-signals