The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill includes instructions to install additional agentic skills from the author's monorepo using a provided management script. Since these originate from the vendor's official repository, they are considered standard functionality.
Evidence: Commands such as python3 scripts/install-external-skills.py --skill codex-loop in SKILL.md.
[COMMAND_EXECUTION]: The skill uses shell commands for setup and installation, including the execution of a Python script and the npx package runner.
Evidence: npx -y skills add ./engineering/skills/agentic-development in README.md and script execution in SKILL.md.
[INDIRECT_PROMPT_INJECTION]: The skill operates by reading repository facts and local rules from the user's workspace, creating a surface where malicious content in those files could influence the agent's behavior.
Ingestion points: SKILL.md instructions for reading "repo facts" and "local skill rules".
Boundary markers: The skill instructs the agent to prioritize local safety gates and constraints.
Capability inventory: The skill can execute local scripts and file operations.
Sanitization: No specific input sanitization or validation process for these ingested files is documented.

agentic-development