The Agent Skills Directory

[COMMAND_EXECUTION]: The instructions direct the agent to run node ui/dist/adapter.mjs --serve in the project terminal. This facilitates the execution of local scripts with environment variables (CRAFT_WORKSPACE_ROOT) provided by the agent's context.
[DYNAMIC_EXECUTION]: The skill relies on a runtime-build and execution workflow for its UI component (ui/dist/adapter.mjs). It instructs the agent to build the component if missing and then execute it, which involves running dynamically prepared local files.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted geometric and drawing data which could potentially contain malicious instructions targeting the agent's broad tool capabilities.
Ingestion points: Drawings loaded via open/new_drawing and entity data retrieved via query_entities (SKILL.md).
Boundary markers: Absent; the instructions do not specify delimiters or tell the agent to ignore instructions embedded in the CAD data.
Capability inventory: Shell execution (node), tool command execution (execute), and filesystem writes (save).
Sanitization: Absent; there is no instruction for the agent to sanitize or validate the content of the entities before processing them.

cadcraft-editor