The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local shell commands such as ls, cat, and grep within SKILL.md to identify the project structure and dependencies. This allows it to tailor recommendations to the specific technologies used in the codebase.
[EXTERNAL_DOWNLOADS]: The skill provides instructions and recommendations for downloading and installing external MCP servers and official plugins, including context7, Playwright, and the GitHub MCP server. These references point to well-known technology providers and services.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes untrusted metadata from the user's project files without sanitization.
Ingestion points: Metadata and dependency lists are read from files like package.json and pyproject.toml in SKILL.md using the cat command.
Boundary markers: No boundary markers or 'ignore' instructions are provided to the agent when processing the content of these files.
Capability inventory: The skill possesses the Bash, Read, Glob, and Grep tools.
Sanitization: There is no evidence of filtering or escaping logic applied to the data ingested from the codebase configuration files.

claude-automation-recommender