Summarizes, extracts, and rewrites content from files.
Walkmark
About Walkmark
Walkmark is an open-source desktop application for macOS, Windows, and Linux that records screen activity and converts it into step-by-step documentation. Users select a screen or window, perform a task while the app automatically captures screenshots at each meaningful interaction, and then use AI to generate guides, onboarding flows, or tutorials. The AI model analyzes the full sequence of screenshots to produce coherent step-by-step instructions with consistent context, avoiding repetition or contradictions. Exports are available in Markdown, HTML, or PDF formats, each including numbered steps and embedded screenshots. The app operates locally by default, with AI writing handled either through local models like Ollama, third-party APIs, or an optional hosted service. Editing tools allow users to reorder steps, redact sensitive fields, or adjust framing before finalizing the output.
Key features
- Automatic screenshot capture on clicks, typing, and scrolling
- AI-generated step-by-step documentation from sequences
- Redaction tools for sensitive fields like passwords
- Local-first operation with offline recording and export
- Support for Ollama, OpenAI, Gemini, and other vision-capable models
- Markdown, HTML, and PDF export formats
- Step reordering and frame adjustment in editor
- Clipboard copy of formatted text
Use cases
- Creating internal onboarding guides for new employees
- Generating customer support documentation from troubleshooting steps
- Producing product how-to guides for software releases
Pros
- Records screen interactions automatically during workflows
- Supports multiple AI models including local and cloud options
- Exports to Markdown, HTML, or PDF with embedded screenshots
- Operates offline for recording and export
- Open-source with MIT license and no subscription required
Cons
- Requires screen capture and accessibility permissions
- Wayland support on Linux may need additional setup
- Walkmark Cloud is not yet available
Frequently asked questions about Walkmark
What is Walkmark?
Walkmark is a free, open-source desktop application for macOS, Windows, and Linux that records screen activity and converts it into AI-generated step-by-step documentation. It captures screenshots during meaningful interactions and uses AI to produce coherent guides, onboarding flows, or tutorials.
How does Walkmark work?
Users select a screen or window, perform a task while the app automatically captures screenshots at each click, type, or scroll. The AI then drafts titles and step-by-step instructions from the sequence, which can be edited and exported in Markdown, HTML, or PDF formats.
Do I need an account to use Walkmark?
No account is required for the open-source desktop app. Walkmark Cloud, a hosted AI writing service, will be optional and may require an account when it launches.
Does data leave my machine when using Walkmark?
Recording, editing, and export remain local by default. Screenshots and prompts only leave your machine when you click 'Write' to use a local model, your own API provider, or Walkmark Cloud when available.
Which AI models work with Walkmark?
Walkmark supports any vision-capable AI model, including local models like Ollama, third-party APIs such as OpenAI, Gemini, or Anthropic, or an OpenAI-compatible local server.
What export formats does Walkmark support?
Walkmark exports guides in Markdown (with an images folder), self-contained HTML, or print-ready PDF. Users can also copy formatted text to the clipboard from the editor.