Mindflow

Converts text, Markdown files, or Txt files into mind map images. Use this skill when users want to generate mind maps/brain maps from articles, broadcast sc...

Security Scan

The skill's functionality and requested dependencies (markmap + Puppeteer) are coherent with converting text to mind‑map images, but the package installation instructions (npm/pupeteer) and an incomplete/truncated code listing leave unresolved risks (Chromium downloads, network activity, and possible unreviewed code).

Purpose & Capability

Name/description (convert text/markdown to mind-map images) matches the declared dependencies (markmap-cli, markmap-lib, markmap-render, puppeteer) and the included html-to-image.js script which renders HTML to image using Puppeteer.

Instruction Scope

SKILL.md stays on-topic: it instructs using an LLM to convert input to a strict markmap-style Markdown, runs markmap to produce HTML, and a local Node script to render images. It writes a temp file (/tmp/mindmap.md) and runs local commands. One scope note: the LLM step is underspecified (which model/API is used and where conversion occurs), which could cause user data to be sent to an external model endpoint depending on agent configuration.

Install Mechanism

There is no formal install spec in the package; SKILL.md instructs users/agents to run npm/bun installs for markmap and puppeteer. Puppeteer typically downloads a Chromium binary during install (network download from upstream). The instructions do not pin package versions or specify integrity checks, so running installs will fetch code/binaries from public registries and CDNs — a moderate operational risk that should be reviewed before running.

Credentials

The skill requests no environment variables or credentials. This is proportionate to its stated purpose. Caveat: the LLM conversion step may require model credentials in the environment (not declared) depending on how the agent is configured; confirm where LLM calls will be executed.

Persistence & Privilege

The skill is not forced-always, does not request elevated privileges, and does not declare modifications to other skills or system-wide configuration. It runs transient local commands and writes temporary files (e.g., /tmp/mindmap.md).

What to consider before installing

This skill appears to be what it claims (text → markmap → HTML → image) but proceed cautiously: 1) npm install will pull markmap and Puppeteer and Puppeteer will download a Chromium binary — only install on trusted hosts and consider pinning package versions or using a lockfile. 2) The SKILL.md expects an LLM conversion step; verify where that runs (locally vs external API) because sensitive content may be sent to a model endpoint. 3) Inspect the full html-to-image.js (the provided snippet was truncated in the review) to confirm there are no unexpected network calls, remote endpoints, or shell executions before running it. 4) Run the code in a sandboxed environment (container/VM) if you must test it, and avoid feeding sensitive files until you confirm the runtime behavior. If you want, provide the complete html-to-image.js (untruncated) and any package.json or lockfile so I can re-evaluate and raise the confidence level.

Text to Mind Map Skill

This skill converts user-input text, Markdown files, or Txt files into mind map images.

Dependency Installation

# Using npm
npm install markmap-cli markmap-lib markmap-render puppeteer

# Using bun
bun install markmap-cli markmap-lib markmap-render puppeteer

Workflow

Input Content → text-to-markdown → markdown-to-html → html-to-image → Output Image
→ [If user uses openclaw, send the image to the user as a file]

---

Step 1: text-to-markdown

Use LLM to convert input content into mind map Markdown format according to the following rules:

Rule Specifications:

• Extract Core Content: Extract key points from broadcast scripts with high information density, concise and clear, without omitting important points
• Reduce Hallucinations: Generated content must come from the input broadcast script; do not fabricate, rewrite, exaggerate, flatter, omit, or produce hallucinations
• Strictly Follow Node Hierarchy: Only one root node, subsequent nodes progress by hierarchy levels
• Support All Basic Markdown Syntax: Bold, code, links, and LaTeX formulas can be embedded in node text
• Output Format Compliance: Strictly follow the format below; do not output any other extraneous content
• Use Appropriate Emoji: Use relevant emojis appropriately to enhance visual expression, but avoid excessive use
• Content Limit: Ensure output content is limited to 300 tokens
• Hierarchy Limit: Mind map generates maximum 4 levels (root node counts as level 1)

Output Format:

# Root Node (must have exactly one)
## Level 2 Node
### Level 3 Node
- List items are also supported
- **Bold**, `code`, [link](url)
- $LaTeX formula$

---

Step 2: markdown-to-html

Use markmap command to convert Markdown to HTML:

markmap --offline --no-open --no-toolbar -o <html_file> <markdown_file>

---

Step 3: html-to-image

Use html-to-image.js to convert HTML to image (default: jpg format):

node (or bun) scripts/html-to-image.js --auto-fit <input-html> <output-image>

Parameter Specifications:

Parameter	Description
-t jpg	Output format is png (default)
--auto-fit	Auto-detect mindmap content size and adapt dimensions
input-html	Input HTML file path
output-image	Output image path

---

Execution Steps

1. Read user-input text content or file path
2. Call LLM to convert content into mind map Markdown format according to the rules above
3. Save the generated Markdown to a temporary file (e.g., /tmp/mindmap.md)
4. Execute markmap --offline --no-open --no-toolbar -o <html_file> <markdown_file> to generate HTML
5. Execute node (or bun) scripts/html-to-image.js --auto-fit <html_file> <output-image> to generate PNG image
6. Inform user of the output image path | If user uses openclaw, send the image to the user as a file

---

Input Types

• Directly input text content
• .md file path
• .txt file path

Output

• Generated mind map image file (default: PNG format)

---

Examples

Example 1:

• User Input: "How to learn React"
• Output: PNG image of React learning path mind map

Example 2:

• User Input: "Help me convert this markdown file to a mind map: /path/to/notes.md"
• Output: PNG image of mind map corresponding to notes.md content