The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and visualize project-specific data, including other skill definitions.
Ingestion points: The script scripts/diagram.mjs reads user-authored workflow.json and layout.json files. The SKILL.md instructions direct the agent to read project documentation and other skill definitions (SKILL.md files) as evidence.
Boundary markers: No explicit delimiters are specified for the ingested content in the instructions.
Capability inventory: The Node.js helper performs file-system write operations to generate diagram.html and temporary files, and it can start a local HTTP server for previews.
Sanitization: All input data for the diagram is strictly validated against JSON schemas before processing, which significantly reduces the risk of accidental execution of embedded instructions.
[DYNAMIC_EXECUTION]: The skill employs dynamic code generation for schema validation.
Evidence: The file assets/model.mjs contains a bundled version of the AJV validation library, which uses the new Function() constructor to compile JSON schemas into executable JavaScript functions at runtime.
[COMMAND_EXECUTION]: The skill requires the execution of a Node.js helper script.
Evidence: Instructions in SKILL.md and README.md guide the agent and user to run node scripts/diagram.mjs for checking, building, and serving diagram previews. The preview server is securely bound to the loopback interface (127.0.0.1).

wkc-workflow-diagram