setup
Pass
Audited by Gen Agent Trust Hub on May 5, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: Utilizes
curlto perform HTTP requests to the vendor's API endpoints atapi.deepread.tech. It also usespython3to parse and format JSON data returned from the API.\n- [DATA_EXFILTRATION]: Reads from and writes to the local.envfile to manage theDEEPREAD_API_KEY. This behavior is transparently documented and serves the legitimate purpose of local credential management for the service.\n- [SAFE]: The skill implements a secure device authorization flow where the user approves the agent's access via a browser, ensuring the API key is handled directly by the agent without requiring the user to copy/paste sensitive credentials into the chat.
Audit Metadata