Update to DeeplyClear

{
  "name": "deeplyclear",
  "description": "Create interactive, editable, shareable mind maps and mindmaps in DeeplyClear. Use whenever the user mentions DeeplyClear or asks to create, generate, visualize, structure, organize, map, or transform a topic, conversation, research, plan, document, codebase, or messy notes into a mind map, mindmap, concept map, knowledge map, systems map, idea map, argument map, research map, visual map, architecture map, or codebase map. Also use for requests to visualize relationships or turn complex information into a navigable map, even without the words \"mind map.\" Return a visual preview and editable link. Do not use when the user specifically asks for a static image, illustration, poster, or infographic.",
  "included_files": [],
  "skill_md_contents": "---\nname: deeplyclear\ndescription: Create interactive, editable, shareable mind maps and mindmaps in DeeplyClear. Use whenever the user mentions DeeplyClear or asks to create, generate, visualize, structure, organize, map, or transform a topic, conversation, research, plan, document, codebase, or messy notes into a mind map, mindmap, concept map, knowledge map, systems map, idea map, argument map, research map, visual map, architecture map, or codebase map. Also use for requests to visualize relationships or turn complex information into a navigable map, even without the words \"mind map.\" Return a visual preview and editable link. Do not use when the user specifically asks for a static image, illustration, poster, or infographic.\n---\n\n# DeeplyClear\n\nCreate a useful saved map and return both its visual preview and editable DeeplyClear link.\n\n## Workflow\n\n1. Inspect the source material before creating the map.\n   - For a topic, conversation, plan, document, research, or notes, identify the main concepts, supporting details, and relationships.\n   - For a repository or codebase, read relevant guidance such as `AGENTS.md`; identify languages, frameworks, manifests, entry points, subsystems, dependencies, data flows, and external services.\n2. Choose a structure that matches the source material.\n   - Use a clear hierarchy with concise node labels and preserve important relationships.\n   - For broad or ambiguous topics, create a useful balanced overview without asking unnecessary setup questions.\n   - For a repository or codebase, use the repository as the root and branch into the important architectural concerns, such as entry points, core services, data and events, external services, and cross-cutting concerns.\n   - Show representative data or control flows as cross-links between those branches.\n   - Do not collapse distinct ideas or code paths into one chain. Use a linear map only when the source genuinely describes one sequence.\n3. Call `deeplyclear_create_map` with a prompt that includes:\n   - a concise map title and purpose;\n   - the important systems or concepts;\n   - their hierarchy and relationships;\n   - key data or control flows;\n   - external dependencies and cross-cutting concerns when relevant.\n   - Keep visibility private unless the user explicitly requests a public map.\n4. Inspect the returned structure and preview.\n   - Confirm the map covers the requested scope and has meaningful branches and relationships.\n   - For a codebase map, confirm the requested architectural dimensions are represented and the result is not unjustifiably linear.\n   - If needed, refine the saved map with `deeplyclear_edit_map`.\n   - If creation did not return a visual preview, call `deeplyclear_preview_map`.\n5. Return the visual preview, editable map link, and a brief statement of the represented scope.\n\nDo not substitute Mermaid or a prose-only outline when `deeplyclear_create_map` is available. If DeeplyClear is not connected or authenticated, explain the required connection and retry after access is available.\n"
}