skills/samhvw8/dot-claude/mise-expert/Gen Agent Trust Hub

mise-expert

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides examples for installing the 'mise' tool in CI/CD and Docker environments using the official installation script from https://mise.run.
  • [COMMAND_EXECUTION]: The skill is centered around defining and running shell-based tasks within mise.toml files, which involves executing arbitrary commands as part of a development workflow.
  • [INDIRECT_PROMPT_INJECTION]: The skill involves reading and processing external configuration files (e.g., mise.toml, .env).
  • Ingestion points: Project configuration files such as mise.toml, .tool-versions, and .env (SKILL.md).
  • Boundary markers: Not explicitly defined in the provided configuration examples.
  • Capability inventory: Execution of shell commands via task runner run properties and environment variable interpolation.
  • Sanitization: No specific sanitization or validation of the content within these configuration files is demonstrated, relying on the user's control over the project environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 04:02 PM
Security Audit — agent-trust-hub — mise-expert