The Agent Skills Directory

[SAFE]: The skill implements strict quality control and security auditing practices that focus on automation and consistent enforcement across local and CI environments.
[COMMAND_EXECUTION]: Recommends standard development CLI tools such as eslint, mypy, pytest, and npm audit for their intended purposes. These commands are configured with hardened flags, such as --max-warnings 0, to ensure strict compliance with quality standards.
[EXTERNAL_DOWNLOADS]: Fetches documentation and configuration guidelines from official and reputable sources, including TypeScript, ESLint, and the RustSec Advisory Database. These references are used to provide the agent with authoritative information on security tooling.
[INDIRECT_PROMPT_INJECTION]: The skill processes project configuration files, which presents a surface for potential instruction injection.
Ingestion points: Reads project manifest files (package.json, pyproject.toml, Makefile, Cargo.toml, go.mod), linter/type-checker configurations, and CI workflow files.
Boundary markers: The skill instructs the agent to inventory existing project scripts but does not explicitly define delimiters for the ingested content.
Capability inventory: Includes shell execution for common ecosystem tools like npm, pytest, tsc, and eslint.
Sanitization: Risk is mitigated by providing predefined command templates for the agent to follow and requiring human review and disclosure for any modifications to the gate configurations.

quality-gates