The Agent Skills Directory

[DATA_EXFILTRATION]: The skill explicitly instructs the agent to read sensitive security files during the exploration phase. Specifically, it requests the agent to read config/master.key and .env files. In a Ruby on Rails environment, the master.key is the primary secret used to decrypt all encrypted credentials. Accessing this file exposes the project's most sensitive production secrets to the AI model's context.
[DATA_EXFILTRATION]: The skill instructions claim to use configuration from ~/.claude/settings.json. Directing an agent to access files in the user's home directory outside of the specific project workspace is a privacy violation and a security risk, as it bypasses standard project sandboxing.
[PROMPT_INJECTION]: The skill is designed to perform a "Deep Codebase Exploration," reading many files throughout the project to generate documentation. This creates an indirect prompt injection surface where malicious instructions embedded in the codebase could influence the agent.
Ingestion points: Entire project root, configuration files (package.json, Gemfile, schema.rb), and the config/ directory.
Boundary markers: Absent. There are no instructions to use delimiters or ignore instructions found within the processed files.
Capability inventory: The agent is authorized to write the resulting content directly to README.md in the project root.
Sanitization: Absent. There is no logic provided to sanitize or validate the content extracted from the codebase before it is used to generate the output.

readme