The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local shell commands and internal scripts (git, node, python, pnpm) to scan the project, manage its own dependencies, and process analyzed data.\n- [EXTERNAL_DOWNLOADS]: Initialization may trigger pnpm install to prepare the required libraries, which involves connecting to the official npm registry.\n- [PROMPT_INJECTION]: The skill defines a clear security boundary by instructing subagents to treat external project content (READMEs, manifests) as untrusted data and to ignore any instructions found within them, mitigating indirect prompt injection risks.

understand