The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses dynamic context injection (the !ls ... pattern) in SKILL.md to list project configuration files during initialization. This is a legitimate use for context-aware tooling.
[COMMAND_EXECUTION]: The skill executes a bundled Python script (detect-linters.py) to scan the filesystem for project metadata. This script uses standard file read operations to identify which linters are applicable to the current project.
[COMMAND_EXECUTION]: The skill facilitates the execution of well-known development tools such as ruff, npx eslint, and composer. These operations are limited to the project directory and are standard for a linting tool.
[REMOTE_CODE_EXECUTION]: While the skill uses npx, which can fetch packages from the npm registry, it is targeting established tools (eslint, tsc, prettier) within a developer workflow, which is consistent with its stated purpose.

lint