The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from external repositories, creating a surface for potential indirect prompt injection attacks.
Ingestion points: Reads project files including README.md, package.json, configuration files, and source code (SKILL.md).
Boundary markers: Not present. The instructions do not specify the use of delimiters or warnings to ignore instructions found within the analyzed files.
Capability inventory: The skill primarily performs file read operations and generates documentation. It includes an instruction to offer to save a summary file (file-write capability).
Sanitization: None specified. The skill processes the files directly to build a mental model.
[SAFE]: The skill follows secure practices for its intended purpose, such as suggesting the scan of .env.example rather than sensitive .env files, and it does not utilize network tools or arbitrary command execution.

codebase-onboarding