diagnose-config

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for the static analysis of MegaLinter configuration files (.mega-linter.yml or .megalinter.yml) to ensure they adhere to expected patterns and hierarchies.
  • [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface as it ingests untrusted configuration data. Ingestion points: local configuration files referenced via $ARGUMENTS or default paths. Boundary markers: Absent. Capability inventory: Local file system access (Read, Grep, Glob). Sanitization: Absent. The risk is minimized by the restricted toolset and the lack of network or code execution capabilities.
  • [COMMAND_EXECUTION]: The instructions guide the agent to validate the syntax of PRE_COMMANDS and POST_COMMANDS within the configuration files. This is limited to diagnostic validation of the configuration content; the skill does not execute these commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 06:55 PM
Security Audit — agent-trust-hub — diagnose-config