Tutorial — RAW Compiler 1.1.4

1. Purchase a License

Visit the purchase page and choose a plan and purchase a license.

2. Receive Your License Key

Within a few minutes of purchase, you'll receive an email with your license key file attached. The file is named something like:

raw-compiler-license-XXXXXXXX.key

Save it to ~/keys/ on your computer. You'll point RAW Compiler to it during configuration.

3. Download the App

Click the link below — the download will start and you'll be brought back to this tutorial. Wait for the download to complete.

The download is a .dmg disk image (~15 MB). It's signed and notarized by Apple — macOS will not block it.

4. Install RAW Compiler

1. Open the downloaded RAW-Compiler-1.1.4.dmg file — it mounts and shows RAW Compiler.app
2. Open a Finder window and navigate to Applications
3. Drag RAW Compiler.app from the mounted disk image into the Applications folder

That's it. No installer wizard, no system extension, no admin password required.

5. Start the App

1. Open Finder → Applications
2. Double-click RAW Compiler
3. The app opens with its main window — you'll see the server status panel and configuration fields

On first launch, macOS may show a Gatekeeper prompt confirming the app is from an identified developer. Click Open to proceed.

6. Confirm the Lightroom Classic Plugin Installed & Enabled

The Lightroom Classic plugin is bundled inside RAW Compiler.app and installs automatically into Lightroom Classic when the app starts. Verify it's enabled:

1. Open Adobe Lightroom Classic
2. Go to File → Plug-in Manager…
3. Find RAW Compiler in the list
4. Confirm its status is Enabled (green indicator). If it shows as Disabled, click Enable to enable it.

7. Activate Your License

The app opens in a stopped state — it won't connect your LLM AI to Lightroom Classic until you configure it and click Start. Before you click Start, set up your license:

License File

You'll receive your license key file by email after purchase. Save it somewhere you won't accidentally delete it — we recommend ~/keys/.

In the app, set License File to the full path of your .key file. When you click Start, the license is verified with the server. It will be periodically re-verified while running — keep an internet connection available. All other settings have sensible defaults.

8. Start Listening for LLM AI Commands

1. In the RAW Compiler app, click the Start button
2. The status indicator turns green — the bridge server is now running
3. The server listens on the ports you configured (defaults: 8765, 8766, 8767)

Make sure Lightroom Classic is open with a photo selected in Develop mode (press D). The plugin will automatically connect to the bridge.

9. Connect an LLM AI

RAW Compiler works with any LLM AI tool that supports MCP (Model Context Protocol) or can make HTTP requests. Pick the section below that matches your LLM AI tool.

{
  "mcpServers": {
    "raw-compiler": {
      "command": "/Applications/RAW Compiler.app/Contents/MacOS/raw-compiler-server",
      "args": ["--stdio"]
    }
  }
}

{
  "mcpServers": {
    "raw-compiler": {
      "command": "/Applications/RAW Compiler.app/Contents/MacOS/raw-compiler-server",
      "args": ["--stdio"]
    }
  }
}

{
  "mcpServers": {
    "raw-compiler": {
      "command": "/Applications/RAW Compiler.app/Contents/MacOS/raw-compiler-server",
      "args": ["--stdio"]
    }
  }
}

{
  "mcpServers": {
    "raw-compiler": {
      "command": "/Applications/RAW Compiler.app/Contents/MacOS/raw-compiler-server",
      "args": ["--stdio"]
    }
  }
}

10. Sample Prompts

Here's a complete editing session you can try right now. Open your LLM AI chat and type these prompts one at a time:

"Show me the current develop settings on the photo"

The AI will read and display all current slider values.

"Brighten the exposure, reduce highlights, lift shadows, and add some vibrance"

Watch Lightroom Classic — the sliders move in real time. The photo updates immediately.

"Add a gentle S-curve with slightly lifted blacks"

The AI creates a custom tone curve for added contrast and mood.

"Get the current crop and set it to 16:9"

The AI reads the current crop and reframes the composition.

"Export a 800x600 JPEG preview to /tmp/landscape.jpg at 95% quality"

Open the exported file to see your result.

"Reset everything back to the original"

Returns the photo to its original state — all edits removed.

More Prompts to Try

• "Apply a moody film look — desaturated, lifted blacks, cool shadows"
• "List my presets and apply one that looks good for portraits"
• "Crop tighter — remove 10% from each side"
• "Make the skin tones more natural"
• "Export a 2000px JPEG and analyze it with Gemini" (Pro only)

11. Sample Workflow

Here's a more advanced workflow that demonstrates the agentic loop — the LLM AI iterates through multiple settings, exports results, and draws a conclusion for you.

The Prompt

"For the current image, set exposure to 0.70, then export to /tmp/image1.jpg. Then set exposure to 0.75, export to /tmp/image2.jpg. Then 0.80, export to /tmp/image3.jpg. Then 0.85, export to /tmp/image4.jpg. After all exports, analyze the images and decide which exposure setting provides the most optimal brightness. Explain your recommendation."

What Happens

The LLM AI will:

1. Set exposure to 0.70 → export /tmp/image1.jpg
2. Set exposure to 0.75 → export /tmp/image2.jpg
3. Set exposure to 0.80 → export /tmp/image3.jpg
4. Set exposure to 0.85 → export /tmp/image4.jpg
5. Analyze all four images (using LLM AI or advanced Gemini APIs if available)
6. Provide a conclusion and recommendation on optimal brightness

This demonstrates the power of agentic workflows — one prompt drives multiple edits, exports, and an AI review pass, all without manual intervention.

12. Advanced: Health Check

To verify the full chain is working, open Terminal and run:

curl http://localhost:8766/health

A healthy response looks like:

{
  "status": "ok",
  "lightroom": {
    "connected": true,
    "version": "15.4",
    "plugin_version": "1.1.4"
  }
}

What each field means:

• "status": "ok" — the bridge is running
• "connected": true — the Lightroom Classic plugin is actively communicating
• "version": "15.4" — your Lightroom Classic version

If the health check fails, verify RAW Compiler is running and Lightroom Classic has the plugin enabled.

Verify Develop Mode

In Lightroom Classic, select a photo and press D to enter Develop mode. Then run:

curl -X POST http://localhost:8766/get_develop_settings

You should get back a JSON object with current settings like Exposure, Contrast, Whites, etc. If the response contains "settings": {} (empty object), you're not in Develop mode — press D.

13. Advanced: Port Configuration

Ports are preconfigured with sensible defaults. You only need to change them if there's a conflict with another app on your machine.

If you change ports:

1. Click Stop to stop serving
2. Set the new port numbers in the app
3. Open Lightroom Classic → Plug-in Manager → select RAW Compiler plugin → set the matching Plugin Port number
4. Update your AGENTS.md references to use the new Client Port and MCP Port values

14. Advanced: Direct HTTP API (Scripts & Custom Agents)

You don't need MCP — any tool that can make HTTP requests can control Lightroom Classic directly via port 8766:

# Health check
curl http://localhost:8766/health

# Get current develop settings
curl -X POST http://localhost:8766/get_develop_settings

# Set exposure and contrast
curl -X POST http://localhost:8766/set_develop_settings \
  -H "Content-Type: application/json" \
  -d '{"settings": {"Exposure": 0.5, "Contrast": 20}}'

# Export a JPEG
curl -X POST http://localhost:8766/export_image \
  -H "Content-Type: application/json" \
  -d '{"folder": "/tmp/output", "filename": "edit.jpg", "width": 2000, "quality": 90}'

This is ideal for Python scripts, shell automation, or building your own agentic loop. The full API reference is in the AGENTS.md file.

15. Advanced: Gemini SA Key File Pro

To enable Gemini AI vision features, download a Service Account JSON key from your Google Cloud project and save it to ~/keys/.

In the app, set Gemini SA Key File to the path of your JSON key file. This is optional — RAW Compiler works without it, but Gemini features will be unavailable.

16. Advanced: Batch Processing Pro Beta

Batch processing allows your LLM AI to edit multiple photos concurrently. This feature is currently in beta — detailed documentation coming soon.