The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides documentation and examples for executing shell commands and managing child processes within a Node.js environment.\n
Evidence: references/node-essentials.md contains detailed examples for using the child_process module, including spawn, exec, and execFile.\n
Evidence: SKILL.md includes a core workflow step requiring the agent to run the eslint --fix command.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external JavaScript source files and configuration data, which could contain malicious instructions designed to exploit the agent's capabilities.\n
Ingestion points: The agent reviews .js, .mjs, .cjs files and package.json configurations provided by the user (as specified in SKILL.md).\n
Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded prompts within the code or comments it analyzes.\n
Capability inventory: The skill utilizes file system operations (fs/promises), network requests (fetch), and command execution (child_process) across its reference guides.\n
Sanitization: There are no explicit instructions to sanitize or validate external content before processing or interpolating it into prompts.\n- [DATA_EXFILTRATION]: The skill documents the use of APIs that can be used to read sensitive local data and transmit it to external endpoints.\n
Evidence: references/node-essentials.md provides examples for using fs/promises to read files and process.env to access environment variables.\n
Evidence: references/browser-apis.md and references/async-patterns.md provide extensive documentation for the fetch API for network communication.

javascript-pro