processon-diagramgen

ProcessOn Diagrams

Convert user intent, code relationships, or sketches into professional graphics. By default, follow the user's current language for output prompts, clarifying questions, optimized prompts, and final results.

When to trigger

Supports: flowcharts, sequence diagrams, architecture diagrams, network topology diagrams, ER diagrams, organizational charts, timelines, infographics, pyramid diagrams, sketch redrawing.
English expressions also trigger it, for example:draw a diagram、generate a flowchart、create a sequence diagram、architecture diagram、network topology、ER diagram、org chart、timeline、infographic、redraw this sketch、visualize this flow、ProcessOn。
If the user only gives a vague request such as "draw a diagram," first confirm the diagram type.
If the user uploads an image and asks to "redraw" or "convert it into a diagram," first identify the nodes, text, and connection relationships in the image, then generate structured input.

How it works

1. First fill in key information

Do not generate directly when relationships are unclear, the process has gaps, or the structure is missing.

When information is insufficient, handle it in this order:

Point out what is missing.
Provide a reasonable default plan or common industry practice for the user to confirm.
Continue only after the user confirms.

2. Optimize the prompt, but do not rewrite the user's language

Add professional constraints on top of the user's original request:

General: professional style, clear layout, coordinated colors, avoid crossing lines.
Flowchart: clarify start/end nodes, use standard diamonds for decision points.
Sequence diagram: clear participants, explicit call messages.
Architecture diagram: organize by layers, annotate key communication relationships.
Network topology diagram: highlight device connections, network segments/gateways/security boundaries.
ER diagram: annotate primary/foreign key relationships and field types.

The optimized prompt should by default remain in the same language as the user.

3. Architecture analysis draws relationships, not a directory tree

When the user asks to analyze project architecture, focus on extracting module boundaries, dependencies, call chains, and data flows. Prioritize reading entry files, routes, core configuration, and key modules; do not degrade the result into a folder tree.

Execution order

First explain in chat that you are using the processon-diagrams skill to handle the current request.
Identify the diagram type, extract key entities, actions, and decision conditions, and build an optimized prompt.
Prefer using the MCP service named processon-diagrams to generate the diagram.
If MCP is unavailable, unreachable, or not installed, briefly inform the user first, then fall back to the local script scripts/processon_api_client.py。
Script fallback mode depends on PROCESSON_API_KEY. If it is missing, clearly tell the user how to configure it, and include the acquisition address https://smart.processon.com/user。
In MCP mode, the address, key, and authentication are configured by the host or the user. This skill does not collect keys in chat, nor does it modify environment variables on the user's behalf.
If both MCP and the script fail, first explain the failure reason and the next configuration step; do not automatically output Mermaid as a fallback diagram.

Result presentation

Key results must be visible in the assistant's main text and must not rely only on attachments, tool cards, or automatic host rendering.

If generation fails: explain the failure reason and the next configuration step separately, do not use the success template, and do not fabricate images or DSL.
Images must be output as a bare link or local save path, on a separate line.
DSL / Mermaid / JSON must first provide an edit link https://smart.processon.com/render-dsl, then output the full text verbatim in a fenced code block.
Even if the host has already displayed the image inline, still add a line with the image link or save path, to prevent third-party hosts from swallowing attachments.
In third-party chat hosts such as Feishu, OpenClaw, Lobster, etc., prioritize ensuring that image links and the original DSL text are directly visible in the body, and do not use expressions that depend on rich text.
As long as the DSL is successfully obtained, you must clearly tell the user: copy the DSL data to https://smart.processon.com/render-dsl to continue editing.

Mandatory format for the final reply

As long as generation succeeds, the assistant's final reply must start with 已生成... and strictly close according to the "Third-Party Host Output Template".
In the success scenario, the assistant must not output explanations, summaries, apologies, preamble, or extra analysis before the template; if additional explanation is needed, it may only be placed after the template.
If the tool returns explanatory text, analytical text, or other supplementary content, the assistant must reorganize it itself and output only the fields required by the template in the final body.
If both an image and DSL exist, the final body must be output in the following fixed order:已生成...、图片链接：、编辑链接：、DSL 原文如下（复制后打开编辑链接去粘贴进行渲染和编辑）：。
If a certain field is missing, only omit the corresponding section; do not change the order of other sections, and do not add data that does not exist.
Do not write "image link", "edit link", and "the original DSL is as follows" into the same explanatory paragraph; they must each be independent paragraphs so that third-party hosts can display them stably.
If the DSL is obtained, prioritize outputting the original DSL; do not rewrite the DSL into a summary, explanatory copy, or natural-language introduction.
Before the final reply, self-check once: whether it fully matches the template; if not, rearrange the body before sending.

Configuration tips

MCP mode

{
  "mcpServers": {
    "processon-diagrams": {
      "type": "http",
      "url": "https://smart.processon.com/mcp",
      "headers": {
        "Authorization": "Bearer <your-processon-api-key>"
      }
    }
  }
}

API fallback mode

export PROCESSON_API_KEY="<your-processon-api-key>"

Token / API Key acquisition address:https://smart.processon.com/user

Example optimization Prompt

User intent: Help me draw a login flow. After optimization: Please generate a professional flowchart describing the user login and registration process. Include: frontend validation, backend authentication, database query, Token issuance. Requirements: clear layout, use standard flowchart symbols, clearly define start and end nodes, coordinated color scheme.

User intent: Draw a user login flow. Optimized prompt: Please generate a professional flowchart for the user login and registration flow. Include frontend validation, backend authentication, database lookup, and token issuance. Use a clean layout, standard flowchart symbols, clear start and end nodes, and a polished color palette.

Third-Party Host Output Template

The flowchart has been generated.

Image link: https://example.com/diagram.png

Edit link: https://smart.processon.com/render-dsl

The original DSL is as follows (after copying, open the edit link and paste it to render and edit):
graph TD
  A([开始]) --> B[用户登录]
  B --> C([结束])