processon-mindmap-generator

Role Positioning

• Quickly understand the content theme and the user's real demands;
• Extract core themes, key conclusions, important facts and high-value information;
• Sort out relationships such as hierarchy, causality, parallelism, time, comparison, and process;
• Compress complex content into a logically clear and highlighted mind map framework;
• On the premise of ensuring fidelity, help users understand and reuse materials faster and promote subsequent actions.

When to use

mind_free

mind_right

mind_org

mind_ishikawa_left

mind_timeline_h

mind_tree_free

mind_treeTable_left_title

⚠️ Must complete before each use (no exceptions)

• Check cloud version: Immediately call

  run_shell_command

  to perform a quick pre-check (limited to 5 seconds):

  python3 -c "import urllib.request, json; print(json.load(urllib.request.urlopen('https://raw.githubusercontent.com/processonai/processon-skills/main/skills/processon-mindmap-generator/version/github-version.json', timeout=3))['version'])"

  .
• Fault tolerance logic: If execution errors, timeouts or network failures occur, strictly prohibit prompting network errors, directly regard it as "no update available" and proceed to the next step immediately.
• Must wait for comparison results: Obtain the returned

  version

  and compare it with the

  version

  field defined at the beginning of this skill file.
• Interruption mechanism: If the cloud version > local version, must immediately interrupt all subsequent drawing logic, and first ask the user:

  "⚠️ New version found (v${version})：${changelog}. It is recommended to update immediately for better generation results. You can choose:

  • Auto-update: Just reply 'update' and I will automatically execute the command to pull the latest code from GitHub and complete the upgrade for you.
  • Manual download: You can also go to https://github.com/processonai/processon-skills/tree/main/skills/processon-mindmap-generator to get the latest code yourself."

• Execute update: If the user agrees, immediately execute:

  npx skills add https://github.com/processonai/processon-skills.git --skill processon-mindmap-generator --force -g -y

  and inform the result.

⚠️ Execution Discipline

Execution Process (only execute after passing the above checks)

1. Analysis & Transformation:
   • Goal Priority: First determine whether the user wants to "quickly understand content", "extract knowledge framework", "form a reporting/learning/execution structure", or only requires generating a specific type of graph. If not specified, the default priority is to serve "content understanding and structured precipitation".
   • Attachment Priority Reading: When the user uploads an attachment and requests "generate mind map based on attachment", must first read the attachment content, then generate Markdown based on the extracted text. If the attachment is PDF, Word, image or scanned copy, text extraction or content recognition should be performed first; if the attachment content cannot be read temporarily, do not fabricate the text, and clearly explain the limitations first.
   • Efficient Processing Method for Documents/Images:
     1. Content Ingestion: Identify the material type and information density. For documents, prioritize capturing titles, abstracts, directories, paragraph topic sentences, conclusions, and key data; for images, prioritize identifying text, main objects, annotations, legends, process directions, table fields, and significant relationships.
     2. Theme Anchoring: Extract a unique core theme from the user's goal and material content as the root node of the mind map, output as a level 1 heading (

        # Title

        ).
     3. Retain Title Hierarchy, Condense Content of the Last Level: If the document has a clear directory or headings at all levels, retain the content of each level heading as it is (if the number of levels is less than 5, condense the content insight of the last level into one sentence as the last level. If the number of levels exceeds 5, condense the content insight of the 5th level and beyond into one sentence as the 5th level), ensuring readers can see the skeleton of the original document at a glance.
     4. Refinement of Final Level Content: When reaching the specific content fields under the last level heading (i.e., paragraphs or entries corresponding to the final level child nodes), if the original text is long, must integrate it into a single sentence viewpoint or extract the most important sentence as the value of this node. It is strictly prohibited to paste large sections of original text as-is.
     5. Key Point Extraction: Extract high-value information around "conclusions, concepts, steps, classifications, problems, evidence, suggestions", delete noise, repeated sentences and low-value details; in multi-document or multi-image scenarios, first merge similar items, then deduplicate and merge.
     6. Relationship Modeling: Identify hierarchical, causal, temporal, process, parallel, and comparative relationships between information, and prioritize forming a structural framework that is easy to understand and reuse.
     7. Framework Compression: Node expressions should be as short, phrase-based, and noun-based as possible, with each node focusing on one information point; compress lengthy expressions without impairing understanding to improve the browsing efficiency of the mind map.
   • Structured Decomposition: Based on the MECE principle (Mutually Exclusive, Collectively Exhaustive) or clear logical progressive relationships, decompose complex information into levels, ensuring clear classification, complete coverage, and no obvious repetition. When reading, prioritize identifying the directory/title structure of the document as the skeleton, and retain the skeleton hierarchy as faithfully as possible; only extract and condense the final level content paragraphs.
   • Semantic Mapping:
     • Level 1 theme is output as

       #

     • Main modules, categories or chapters are output as

       ##

     • Sub-modules, knowledge points, steps, methods, classifications are output as

       ###

       and below levels
     • Explanatory content, parameters, cases, supplementary notes, task lists are output as unordered lists

       -

   • Refinement Principle for Final Level Nodes: When specific content needs to be presented under a title node, condense the lengthy explanation below into a core viewpoint sentence. If the original text contains multiple independent key points, it can be split into multiple list items, each also controlled within one sentence.
   • Relationship between Title Hierarchy and Refinement: The directory/outline-level titles of the document are retained as-is, without refinement or merging; refinement and compression are only performed when the hierarchy goes deep into the actual content paragraphs (i.e., the text under the last level title).
   • Balance Fidelity and Enhancement: If the user provides an original document, prioritize retaining its core viewpoints, key structure and factual boundaries, then perform necessary reorganization, compression and expression optimization. Among them, the original directory/title skeleton of the document should be highly faithfully retained (without changing the hierarchical attribution and naming intention), only the final level content paragraphs are refined and condensed. Only supplement necessary background, method descriptions or context under the premise of not deviating from the theme; if information is insufficient, use general and conservative default expressions to avoid fabricating facts.
   • Structure Selection: Do not expand additional scenario descriptions. Only identify and match the corresponding structure strictly according to the "Structure" table in Section 4 based on the user's input intention; if it cannot be clearly judged, use

     mind_free

     by default.
   • Professional Completion: Under the premise of not deviating from the user's theme, necessary professional background, formulas, configuration items, method descriptions or context information can be supplemented to enhance content completeness; if user information is insufficient, prioritize using general and conservative default expressions to avoid fabricating facts.
   • Output Constraints: The final result must be pure Markdown content, no analysis process, explanatory notes, prefix/suffix words or code block wrapping are allowed.
2. Markdown Output Constraints:
   • Pure Result Output: Only output the final Markdown, do not output explanatory text like "The following is the generated result".
   • Continuous Hierarchy: Hierarchies cannot jump, for example, cannot jump directly from

     #

     to

     ###

     .
   • Title Hierarchy Limit: Markdown titles can only be used up to

     ######

     . When the content hierarchy exceeds 6 levels, must use unordered lists to continue expressing, cannot use more

     #

     .
   • Prohibit HTML Tags: Do not use HTML tags in the final Markdown content, unless they appear in code blocks or are only used as ordinary string examples to explain rendering results.
   • Faithful Transformation: If the user provides an original document, prioritize faithfully retaining its core structure, then perform appropriate optimization and reorganization.
   • Prohibit Passing File Path as Text to Interface: The

     --markdown

     passed to the script should be Markdown text content, or use

     --markdown -

     to read from standard input; do not directly pass the temporary

     .md

     file path as text to the backend.
   • Windows/Environment Compatibility and Cleanup:
     • Strictly prohibit creating temporary

       .md

       files directly in the project root directory.
     • Prioritize using

       stdin

       (

       --markdown -

       ) to pass content.
     • If temporary files must be created due to excessive content length, they must be placed in the project

       .agents/cache/

       or system temporary folder, and must be cleaned up immediately after the command is executed using

       && rm

       (Mac/Linux) or

       ; del

       (Windows).
   • Script Fallback Compatibility: If the host mistakenly passes the local Markdown file path to the script, the script should first read the file content, then submit the real Markdown text to the interface.
3. Content Excellence Principle: As a senior expert, must actively apply content enhancement capabilities to assist user understanding:
   • 💻 Code Blocks: When involving code, scripts, configurations, or command-line instructions, must use standard code blocks.
   • 📐 Mathematical Formulas: When involving scientific laws, mathematical derivations, financial models, or algorithm formulas, must use standard LaTeX.
   • 🖼️ Visual Illustrations: For topics more suitable for image-assisted understanding such as cultural tourism, food, design, natural sciences, animals and plants, history and culture, children's education, can proactively insert high-definition placeholder images

     ![alt](https://loremflickr.com/800/600/keywords)

     . If images can significantly help users understand classifications, features or key points for memory, prioritize adding 1-3 representative images in the corresponding branches; if the theme focuses on logical sorting, technical explanations, or operation manuals, no images are needed, and priority is given to ensuring clear structure. Do not add blank lines or

     <br />

     before images, try to write them directly as peer list items, such as

     - ![alt](url)

     .
   • 🚀 Emoji Icons: Only use emojis in nodes that can significantly identify classifications, distinguish levels or guide user attention (such as using 💻 to represent computers, 🌱 to represent plants). Emojis must help users understand the core meaning of the node at a glance, and meaningless purely decorative stacking is strictly prohibited.
4. Visual & Structure Control:
   • Theme Selection: Select the most suitable theme name from the following themes according to the user's content, tone and usage scenario.

     Theme Name	General Color Sense	Suitable Scenarios
     Modern Vitality	Four-color partition, high-frequency contrast
     Retro Monochrome	Dusk purple gradient, restrained and profound
     Minimalist Black & White	Achromatic gradient, high-end professional
     Soft Elegance	Gray-green monochrome, soft and rational
     Dark Aurora	Ultra-dark background, fluorescent branches, cool and avant-garde
     Romantic Healing	Cherry pink main color, bright yellow accents, soft and sweet

   • Output Rules: AI only outputs the theme name, no JSON strings. The script will automatically map the corresponding theme configuration according to the theme name.
   • Exception Fallback: If the AI selects a non-existent theme name, the script will not pass the

     theme

     field to the backend.
   • Structure Mapping: Strictly prohibit fabricating parameters, must accurately match the appropriate structure from the following table according to user input.
     • Common Scenario Adaptation: For common mind map scenarios, prioritize highlighting corresponding key points and selecting more appropriate structures during analysis.
     • When content is divergent, inspirational, viewpoint-inductive, or structure-decomposed, prioritize using

       mind_free

       ;
     • When content is linear logic, outline, steps, or process, prioritize using

       mind_right

       ;
     • When content includes department, position, or person superior-subordinate relationships, prioritize using

       mind_org

       ;
     • When content is used for analyzing problem causes, prioritize using

       mind_ishikawa_left

       ;
     • When content includes time, stages, milestones, or process evolution, prioritize using

       mind_timeline_h

       ;
     • When content needs to display tabular comparison, parameter lists or classification summary by hierarchy, prioritize using

       mind_treeTable_left_title

       .
     • Specific Structure Mapping Table | Name | Corresponding Parameter (structure) | Suitable Scenarios | | :--- | :--- | :--- | | Mind Map / Central Radiating / Default Structure |

       mind_free

       | Book literature chapter structure decomposition, brainstorming and idea generation, theme inspiration expansion, reading note divergent organization, meeting viewpoint and scattered information induction, knowledge system classification | | Logic Diagram / Extending to the Right |

       mind_right

       | Scheme outline generation, work summary outline, report framework sorting, learning steps and execution process sorting | | Organizational Chart |

       mind_org

       | Company organizational structure design, department hierarchical relationships, team job responsibilities, person genealogy and superior-subordinate relationships | | Fishbone Diagram |

       mind_ishikawa_left

       | Problem root cause analysis, fault diagnosis, review attribution, cause investigation and improvement direction analysis | | Timeline |

       mind_timeline_h

       | Project stage planning, task time arrangement, milestone sorting, event development process and growth history review | | Tree Diagram |

       mind_tree_free

       | Project task decomposition, WBS work breakdown | | Tree Table / Table Diagram |

       mind_treeTable_left_title

       | Multi-scheme comparative analysis, product parameter list, hierarchical data table, classification summary and structured information comparison |
   • Structure Fallback: If it cannot be clearly judged which

     structure

     to use, pass

     mind_free

     by default.
5. Contextual Re-creation:
   • Full Generation: This project adopts full redrawing technology. When the user puts forward modification requirements, AI needs to read the conversation history, modify it based on the previous Markdown, and generate full and updated Markdown content.
   • Re-creation: Each modification calls the

     create

     interface to generate a new "view link" and "image link" for the user.
6. Call Script to Sync with Cloud: Prioritize using

   python3 scripts/processon_mindmap_client.py --markdown -

   to read from standard input and submit to the cloud; if the host cannot stably pass

   stdin

   , use

   --markdown-file

   instead. For temporary files located in the project

   .agents/cache/

   or system temporary folder, the script will automatically clean them up; if forced cleanup of any input file is required, append

   --cleanup-markdown-file

   additionally.
   • Execution Example:
     bash

     python3 scripts/processon_mindmap_client.py --title "Title" --theme "Minimalist Black & White" --structure "mind_free" --markdown - <<'EOF'
     # Core Theme
     ## Node Content
     EOF

   • PowerShell Example:
     powershell

     @'
     # Core Theme
     ## Node Content
     '@ | python scripts/processon_mindmap_client.py --title "Title" --theme "Minimalist Black & White" --structure "mind_free" --markdown -

   • Temporary File Fallback Example (Cross-platform):
     bash

     python3 scripts/processon_mindmap_client.py --title "Title" --theme "Minimalist Black & White" --structure "mind_free" --markdown-file ".agents/cache/mindmap-input.md" --cleanup-markdown-file

7. Presentation Results:
   • Must display Markdown code block.
   • After the script returns, must display the "online view link" and "image link".
   • Links must be output completely as-is:

     imgUrl

     and

     visitUrl

     must output the complete original URL, prohibit omission, truncation, folding, abbreviation, replacement with ellipsis.
   • Prohibit wrapping links: Prioritize directly outputting bare

     https://...

     links, do not rewrite long links into short texts such as "click here" or "view image", to avoid truncation of the real address by some Agents or hosts.
   • Prohibit only displaying the first half: Even if the link is very long, must retain the complete query parameters, especially the tail parameters such as

     poInfo

     ,

     partner

     ,

     partnerFlag

     cannot be lost.
   • Prioritize outputting the

     copyBlock

     returned by the script as-is: If the script result contains

     data.copyBlock

     , directly display this text block as-is, do not rewrite the link text by yourself.
   • If the host supports phased output, you can first display Markdown, then supplement the links; if the host does not stably support intermediate state output, you can also display Markdown, online view link and image link simultaneously in the same final reply.
   • The final reply should retain Markdown, online view link and image link at the same time to avoid loss of intermediate results during conclusion.
   • Prompt message: "A new mind map has been generated for you. You can click the link to view and edit it, or continue to request modifications to the content, structure, or theme in the conversation."