The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Node.js scripts (render-company.mjs and validate-company.mjs) to automate document rendering and validation. These scripts use standard Node.js built-in modules and do not initiate network connections or execute arbitrary shell commands.
[PROMPT_INJECTION]: The skill processes untrusted data from local project files (MVP.md, Company.md) using HTML comment boundary markers. While it has file-write capabilities and lacks explicit sanitization, the risk of indirect prompt injection is low as the scripts are limited to structural merging based on specific section IDs. Ingestion points include the local project files, boundary markers are present in the form of HTML comments, capabilities include local file writing, and no specific sanitization is applied to the merged content.
[SAFE]: No network access, hardcoded credentials, or external downloads were detected. The skill follows safe practices for local documentation management and operates within the designated project directory.

mvpfy-document