The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from the local repository (e.g., file names, lockfiles, and manifests) to detect the stack and audit existing health files. This content is processed by the agent to guide the setup process.\n
Ingestion points: scripts/detect-stack.mjs and scripts/scan-repo.mjs scan the local file system for marker files.\n
Boundary markers: Findings from the repository scan are reported to the agent context without specific prompt delimiters.\n
Capability inventory: The skill has the capability to write files via scripts/generate.mjs and modify repository settings using the gh CLI.\n
Sanitization: Strict instructions are provided to the agent to obtain user confirmation via the ask_user tool before performing any writes, commits, or metadata updates.\n- [EXTERNAL_DOWNLOADS]: The skill retrieves configuration templates from well-known and trusted external sources, including the gitignore.io service by Toptal and the official GitHub Licenses API.\n- [COMMAND_EXECUTION]: The skill utilizes the gh CLI and git tools to perform repository audits and apply approved metadata changes, always operating under user oversight.

repo-ready