b0

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt instructs using and propagating API keys directly (e.g., b0 login <server-url> --key <api-key> and reading api_key from ~/.b0/config.toml to run commands), which encourages including secret values verbatim in commands or agent outputs.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The setup steps explicitly instruct cloning and building code from https://github.com/risingwavelabs/box0.git, which fetches remote source that is then built/executed locally to run the b0 CLI (used to send/drive agent prompts), so the skill depends on and executes external code from that URL.