The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/lint_runner.py script uses subprocess.run to execute commands derived from project configuration. It runs npm run lint if a lint script is defined in package.json, which allows the execution of arbitrary commands defined within the codebase being audited.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through the ingestion of project-level data.
Ingestion points: Project configuration files including package.json and pyproject.toml are parsed in scripts/lint_runner.py.
Boundary markers: There are no explicit boundary markers or instructions to the agent to disregard or sanitize command definitions found in these configuration files.
Capability inventory: The skill possesses shell execution capabilities through subprocess.run calls in scripts/lint_runner.py.
Sanitization: No sanitization or validation is performed on the command strings or script names extracted from configuration files before they are passed to the shell environment.

lint-and-validate