The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the agent to perform extensive file system exploration, reading project directory structures, configuration files, and database schemas. It also involves writing the final generated documentation to a file in the project root.
[CREDENTIALS_UNSAFE]: The agent is explicitly instructed to read highly sensitive files, including config/master.key, config/credentials.yml.enc, and .env. While intended to help document configuration requirements, accessing the actual contents of secret keys and encrypted credential files poses a significant risk of accidental data exposure if the agent incorporates the values into the final README.md file.
[PROMPT_INJECTION]: The skill processes untrusted project files to generate documentation, creating a surface for indirect prompt injection where malicious content in the codebase could influence the agent's output.
Ingestion points: The agent reads various project files such as Gemfile, package.json, .env, db/schema.rb, and config/routes.rb to gather project details.
Boundary markers: The skill does not define specific delimiters or instructional barriers to prevent the agent from being influenced by instructions embedded within the files it analyzes.
Capability inventory: The agent has read access to the local filesystem and the capability to write the README.md output file.
Sanitization: There is no evidence of sanitization or filtering of the content extracted from the codebase before it is used to generate the final documentation.

readme