cohere-common-errors

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes diagnostic shell commands using curl to interact with official Cohere status and API endpoints (status.cohere.com, api.cohere.com). These are well-known services and the commands are used legitimately for troubleshooting connectivity and service availability.
  • [DATA_EXFILTRATION]: While the skill utilizes the CO_API_KEY environment variable for authentication tests, it does so by communicating directly with official Cohere API endpoints. There is no evidence of credential harvesting or sending sensitive data to unauthorized third-party domains.
  • [EXTERNAL_DOWNLOADS]: The skill references the standard cohere-ai library for Node.js and the cohere package for Python. It provides instructions to check for their installation using standard package managers (npm, pip).
  • [SAFE]: The skill follows security best practices by recommending the use of environment variables for secret management (e.g., CO_API_KEY) rather than hardcoding credentials.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 01:14 AM
Security Audit — agent-trust-hub — cohere-common-errors