The Agent Skills Directory

[SAFE]: The skill provides a structured framework for workflow analysis based on established UX operating rules and experience core principles.\n- [COMMAND_EXECUTION]: The skill includes a Node.js script, scripts/workflow-ledger.mjs, used for processing workflow data. The script reads local JSON files and calculates step counts and friction tags. It uses standard Node.js modules, does not perform network operations, and contains no dangerous execution patterns like eval or exec.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied JSON files defining workflows. This data ingestion is handled by a script that validates step types and attributes. While it represents an attack surface for untrusted data, the risk is negligible as it is restricted to UX metrics analysis within the agent's context.

workflow-compression