The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted content from various sources, creating a surface for indirect prompt injection.
Ingestion points: Reads local codebase files (README.md, package.json, docs/vision.md, docs/brand-system.md) and external tool outputs (GitHub API, App Store Search API, and package registry data).
Boundary markers: Instructions lack specific delimiters or instructions to ignore embedded commands within the ingested data.
Capability inventory: The skill has the ability to execute shell commands (gh, npm, pip, cargo, gem, dig) and perform network requests (curl).
Sanitization: No sanitization or filtering of external content is mentioned before the data is processed by the AI.
[COMMAND_EXECUTION]: The instructions direct the agent to execute several CLI tools for name availability validation. This includes gh api and gh search for GitHub, npm view for Node.js, pip index for Python, cargo search for Rust, gem list for Ruby, and dig for DNS lookups. These are standard tools for the skill's stated purpose.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch data from well-known services. It queries the Apple iTunes Search API (itunes.apple.com) via curl to check for mobile app collisions and performs HTTP/DNS checks on candidate domains.

nomen