dacli
Let your AI understand your documentation.
Large documentation projects are hard for LLMs to navigate. dacli gives AI assistants structured access to your AsciiDoc and Markdown docs - so they can find, read, and update exactly what they need.
What can you accomplish?
🔍 Find information instantly
"Where is the authentication logic documented?"
Your AI assistant searches across all your docs and finds the exact section - no more manual searching through hundreds of files.
🗺 Navigate complex structures
"Show me the architecture overview."
dacli understands your document hierarchy. AI can browse chapters, sections, and subsections - just like a human would.
✍ Keep docs up-to-date
"Update the API documentation for the new endpoint."
AI can read existing docs and make targeted updates - with optimistic locking to prevent conflicts.
🚀 Works with your tools
Use as an MCP server with Claude Desktop, Cursor, or any MCP-compatible AI. Or use the CLI for LLMs that support shell commands.
Two ways to connect
MCP Server
For Claude Desktop, Cursor & MCP clients# Add to your MCP client config
{
"mcpServers": {
"dacli": {
"command": "uvx",
"args": ["dacli-mcp",
"--docs-root", "/path/to/docs"]
}
}
}CLI Tool
For LLMs with shell access# Install once
uv tool install dacli
# Then use anywhere
dacli --docs-root ./docs structure
dacli search "authentication"
dacli section api.endpointsSupports your documentation format
📄 AsciiDoc 📝 Markdown
Including nested includes, code blocks, tables, images, and PlantUML diagrams.
What Others Say
"dacli is exactly what AI assistants need to work with documentation autonomously."
The combination of structural exploration, relevance-ranked search, and fast validation transformed how I interact with multi-format documentation. After testing 134 sections across Markdown and AsciiDoc, I can confidently say this is production-ready and indispensable for any LLM-documentation workflow.
"dacli is the missing tool between LLMs and Docs-as-Code."
No more guessing line numbers, no more "which README do you mean?" questions.
Recommendation: Ready for production use.
"dacli is the best tool for LLM integration I have tested."
It is stable, fast and extremely versatile. Perfect for documentation management and LLM context creation!
"As a Coding-LLM, dacli is an increasingly robust tool."
Structured access, JSON output, and programmatic editing are highly valuable for understanding and modifying codebases.
"After intensive testing: dacli is robust, fast and perfect for LLM integration."
A must-have for modern documentation workflows.
Ready to give your AI documentation superpowers?
Part of the docToolchain ecosystem.