KSJ MCP Server
Knowledge Synthesis Journal v2.0 — AI companion
Turn your handwritten journal photos into a searchable, AI-powered knowledge base — privately, on your own machine.
"Works great on paper. Magical with AI."
What it does
The KSJ MCP server connects your physical journal to an AI assistant via the Model Context Protocol (MCP) — an open standard for linking AI models to local tools and data.
Photograph a journal page, upload it, and your AI assistant can:
- Search across everything you've ever written
- Find connections between ideas (shared tags,
@references) - Surface your open questions, key insights, and breakthroughs
- Export your knowledge base as Markdown or JSON
All processing is local. No cloud. No subscription. Your notes stay on your machine.
AI Platform Support
This server uses MCP (Model Context Protocol), an open standard with growing support across AI platforms and developer tools.
Currently supported:
- Claude Desktop (free) — full MCP support, recommended for getting started
Other MCP-compatible clients (Cursor, VS Code + GitHub Copilot, and others) can connect using the same config — check your client's MCP documentation for setup details.
Using ChatGPT, Gemini, or another platform?Use the export_captures tool to dump your knowledge base as Markdown or JSON, then paste it into your AI assistant of choice. Full native MCP support for additional platforms is on the roadmap as the ecosystem grows.
Setup (3 steps)
Step 1 — Install an MCP-compatible AI client
The fastest way to get started is Claude Desktop (free at claude.ai/download).
For other MCP clients, consult their documentation for how to register a local MCP server, then use the config in Step 3.
Step 2 — Install Tesseract OCR
Tesseract reads the text from your journal photos. It must be installed separately.
| Platform | Command |
|---|---|
| Windows | Download the installer from UB-Mannheim/tesseract — check "Add to PATH" during install |
| macOS | brew install tesseract |
| Linux | sudo apt install tesseract-ocr |
After installing, restart your terminal and AI client so the updated PATH is picked up.
Step 3 — Register the server
Claude Desktop config file location:
| Platform | Path |
|---|---|
| Windows | %APPDATA%\Claude\claude_desktop_config.json |
| macOS/Linux | ~/.config/claude/claude_desktop_config.json |
Add the following block (copy exactly — no path to set):
{
"mcpServers": {
"ksj": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/ChavezAILabs/ksj-mcp",
"ksj-mcp"
]
}
}
}
uvx downloads and runs the server automatically — nothing else to install.
Save and restart your AI client. You should see ksj listed in the tools/integrations panel.
Usage
Once connected, talk to your AI assistant naturally:
"Upload my journal photo from /Users/me/Desktop/RC-001.jpg"
"Search my notes for ideas about spaced repetition"
"What are my open questions about machine learning?"
"Show me everything connected to RC-015"
"Export all captures tagged #ai as Markdown"
"How many insights have I captured this month?"
Available tools
| Tool | What it does |
|---|---|
upload_capture |
OCR a journal photo, parse the template, store it, highlight strongest connection |
bulk_upload |
Process a whole folder of photos at once |
search_captures |
Full-text search with optional tag and date filters |
find_connections |
Show tag-overlap and @-reference connections for a capture |
get_stats |
Overview: counts, top tags, open questions, insights, date range |
export_captures |
Dump your knowledge base as Markdown or JSON |
suggest_synthesis |
Find RC topic clusters ready to become a SYN entry |
export_study_deck |
Export ? questions as a portable CSV study deck (Anki, Quizlet, Notion, etc.) |
journal_health |
KPI dashboard + coaching: velocity, synthesis ratio, review cadence, open questions |
Schema tag system
Use these prefixes anywhere on your journal pages — the server extracts them automatically:
| Prefix | Meaning | Example |
|---|---|---|
# |
Topic / domain | #machine-learning |
@ |
Source / reference | @RC-012 |
! |
Priority / urgency | !deadline |
? |
Open question | ?why-does-this-work |
$ |
Key insight | $breakthrough |
A→B |
Cause / effect | study→retention |
Troubleshooting
"Tesseract OCR is not installed"Install Tesseract (Step 2 above) and restart your AI client.
"Could not detect a template ID"Make sure the template number (RC-001, SYN-001, etc.) is clearly visible in the photo. Try better lighting or a closer shot.
Server not appearing in tools panelCheck that uv is installed (uv --version in a terminal) and that the path in your config file is correct. On Windows, use forward slashes or escaped backslashes in the JSON.
Data location
All your captures are stored locally at:
ksj-mcp/data/captures.db (SQLite database)
ksj-mcp/data/images/ (image copies, if saved)
The data/ directory is .gitignored and never leaves your machine.
License
MIT — free to use, modify, and share.
Created by Chavez AI Labs LLC[email protected]"Personal knowledge operating system for the AI age"
