Skills
Navigate
skills/manaflow-ai/cmux/cmux-backend/Gen Agent Trust Hub

cmux-backend

Warn

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill directs the agent to access sensitive credential files located in the user's home directory which contain production and development API keys.
  • Evidence: Instructions in SKILL.md specify reading from ~/.secrets/cmux.env, ~/.secrets/cmuxterm-dev.env, and ~/.secret/cmuxterm.env to obtain E2B_API_KEY and FREESTYLE_API_KEY.
  • [COMMAND_EXECUTION]: The skill provides specific shell commands for the agent to execute, including sourcing sensitive environment files and running database migrations.
  • Evidence: SKILL.md contains a bash snippet to source ~/.secrets/cmux.env, and both SKILL.md and references/cloud-vm-control-plane.md instruct the agent to run bun db:migrate:aws-rds-iam.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 27, 2026, 09:58 PM
Security Audit — agent-trust-hub — cmux-backend