agentpmt-docs-and-content
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill utilizes official platform tools like npx and openclaw to install dependencies from the AgentPMT GitHub organization. These are recognized as safe vendor resources.
- [COMMAND_EXECUTION]: Shell commands provided in the documentation are for manual user configuration and setup. There is no evidence of automated or unauthorized command execution by the agent.
- [DATA_EXFILTRATION]: Network activity is restricted to the vendor's official domain (api.agentpmt.com) for retrieving search results, which is consistent with the skill's intended functionality.
- [PROMPT_INJECTION]: The skill ingests search queries through the search actions defined in schema.md. While this is an ingestion surface for external data without explicit boundary markers in the prompt, the capability inventory is limited to vendor API calls and results are sanitized, making the risk inherent to the tool's primary purpose and consistent with safe usage.
Audit Metadata