keyword-signals
Pass
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill defines standard workflows for retrieving keyword trends and market signals via API endpoints. No hardcoded credentials, unauthorized network destinations, or persistent access mechanisms were identified.
- [PROMPT_INJECTION]: The skill was evaluated for indirect prompt injection vulnerability surfaces (Category 8). 1. Ingestion points: External data is fetched from /api/v1/trends/signals and /api/v1/keywords/{slug} endpoints. 2. Boundary markers: Data is ingested as structured API responses. 3. Capability inventory: The agent is limited to network read operations (HTTP GET) and reporting results; no shell command execution, file system writes, or administrative tools are enabled. 4. Sanitization: Not explicitly defined in the provided instructions, though the risk is mitigated by the restricted action scope.
Audit Metadata