The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references and installs external resources from the author and reputable security vendors.
Instructions in AGENTS.md fetch rule configurations from the author's GitHub repository (metyatech/agent-rules).
Installation instructions utilize npx to add the skill from the author's package registry.
SKILL.md recommends the integration of standard, well-known security tools such as osv-scanner, trivy, govulncheck, and OWASP Dependency-Check.
[REMOTE_CODE_EXECUTION]: The skill facilitates the execution of a specific maintenance tool via the package manager.
AGENTS.md instructs the agent to run npx compose-agentsmd or npm install -g compose-agentsmd to manage rule updates. This is a vendor-owned tool designed for the specific purpose of regenerating instruction files.
Safety controls are explicitly included, requiring the agent to present an ANSI-colored diff of planned rule changes and obtain explicit user approval before execution.
[SAFE]: The skill aligns with and enforces industry security best practices.
It mandates the configuration of secret scanning and explicitly forbids the committal of secrets to the repository.
It requires the setup of Software Composition Analysis (SCA) and CodeQL for deep security analysis.
The instructions prioritize 'strict mode' and 'warnings-as-errors' for various compilers and linters.

code-quality-setup