mise
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFE
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill documents the standard installation process for mise using 'curl https://mise.run | sh', which involves downloading and running a script from the project's official domain. It also notes support for the 'asdf:' backend which allows for tool-specific plugin code execution.\n- [COMMAND_EXECUTION]: The skill provides instructions for using 'mise exec' and 'mise run' to execute binaries and tasks, which is the primary intended functionality of the tool manager for development environments.\n- [EXTERNAL_DOWNLOADS]: The skill references the retrieval of tool binaries from official sources such as GitHub, Aqua, and various language-specific package registries (npm, pipx, etc.).\n- [PROMPT_INJECTION]: The skill instructs the agent to fetch external documentation from 'https://mise.jdx.dev' for additional details.
- Ingestion points: Documentation pages from mise.jdx.dev;
- Boundary markers: None;
- Capability inventory: Local command execution and environment management through the mise CLI;
- Sanitization: None.\n- [CREDENTIALS_UNSAFE]: The skill promotes secure secret management by explicitly advising users to load secrets from '.env' files rather than committing them to the main mise.toml configuration file.
Audit Metadata