The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the ecc-plan-canvas CLI utility (associated with the ecc-universal package) to manage review sessions. It executes shell commands for opening artifacts, polling for user feedback, and sending replies.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from local project files and human-in-the-loop feedback, which could potentially contain malicious instructions. 1. Ingestion points: Local project artifacts (.plan.md, .html) and the JSON output from the await feedback loop. 2. Boundary markers: No explicit markers or instructions are provided to separate artifact data from agent commands. 3. Capability inventory: Execution of local CLI tools and the ability to modify project files. 4. Sanitization: The skill does not mention specific sanitization or validation steps for content loaded into the canvas or feedback received from it.
[EXTERNAL_DOWNLOADS]: The skill references a pinned CDN to load Mermaid diagram rendering libraries in the browser and includes a reference to an external GitHub repository as a design inspiration.

plan-canvas