Docs · Two minutes to install

Getting started.

Download CodeFire, add your API key, and connect your AI coding CLI to the 126-tool MCP server.

Heads up: macOS builds are Apple Silicon only. Linux ships x86_64 and arm64. Windows is coming soon.

1

Install CodeFire

Grab the latest signed build from GitHub Releases — free, MIT licensed, no account required.

macOS (Apple Silicon)

  1. 1. Download the latest DMG from GitHub Releases.
  2. 2. Open the DMG and drag CodeFire.app to your Applications folder.
  3. 3. Open CodeFire. Builds are signed and notarized, so it launches without any Gatekeeper warnings.

Requires an Apple Silicon Mac (M1 or later). Data is stored at ~/Library/Application Support/CodeFire/. Updates are checked against GitHub Releases — never silently installed.

2

Open Settings

Configure your API keys and preferences.

Press Cmd+, (or Ctrl+, on Linux), or open the Settings drawer from the tool rail.

Settings covers general preferences, the context Engine, the MCP server, agent defaults and auto-run, Gmail, briefings, and team sync.

3

Add your OpenRouter API key

Required for AI chat, semantic code search, and image generation.

  1. 1. Go to openrouter.ai/keys and create a new API key. Copy it.
  2. 2. In CodeFire Settings, go to the Engine tab.
  3. 3. Paste your API key into the OpenRouter API Key field. It starts with sk-or-v1-...

What this unlocks: The OpenRouter key powers the built-in AI Chat (ask questions about your codebase), Semantic Code Search (search by meaning, not just text), and AI Image Generation. All models are served through OpenRouter with a single key.

Optional settings on this tab:

  • Embedding model — Choose between text-embedding-3-small (faster, cheaper) or text-embedding-3-large (more accurate)
  • MCP server auto-start — Enabled by default. Automatically detects when an AI CLI connects via MCP.
  • Instruction injection — Enabled by default. Adds a managed section to your project's CLAUDE.md telling Claude Code about available CodeFire tools.
4

Connect your AI coding CLI

Install the CodeFire MCP server so your CLI has access to persistent memory and tools.

One-click install

Click the button for your CLI tool. CodeFire will automatically write the MCP config to the right file.

These deep links open CodeFire, which writes the config for you. After installing, restart your CLI session for the connection to activate.

Manual setup

If the one-click install doesn't work, you can configure the MCP server manually.

Finding the MCP server path:

  • macOS: /Applications/CodeFire.app/Contents/Resources/mcp-server/server.js
  • Linux (AppImage): ~/.local/share/CodeFire/mcp-server/server.js (auto-synced on first launch)
  • Linux (deb): /opt/CodeFire/resources/mcp-server/server.js
Claude Code Terminal command
$ claude mcp add codefire node /path/to/resources/mcp-server/server.js

Replace the path with the actual server.js path from the table above.

Codex ~/.codex/config.toml
[mcp_servers.codefire]
command = "node"
args = ["/path/to/resources/mcp-server/server.js"]
Gemini ~/.gemini/settings.json
{
  "mcpServers": {
    "codefire": {
      "command": "node",
      "args": ["/path/to/resources/mcp-server/server.js"]
    }
  }
}
Kimi Code ~/.kimi-code/mcp.json
{
  "mcpServers": {
    "codefire": {
      "command": "node",
      "args": ["/path/to/resources/mcp-server/server.js"]
    }
  }
}
CodeFire Agent opencode.json (project root)
{
  "mcpServers": {
    "codefire": {
      "type": "local",
      "command": ["node", "/path/to/resources/mcp-server/server.js"]
    }
  }
}
5

Add system instructions

Teach your AI agent how to use CodeFire's tools effectively.

Each CLI has a system instruction file where you can add persistent instructions. Copy the CodeFire instructions below into the appropriate file so your agent knows how to use tasks, notes, browser automation, and more. (CodeFire can also inject these automatically — see "Instruction injection" in step 3.)

Claude Code ~/.claude/CLAUDE.md

Add the following to your ~/.claude/CLAUDE.md file (global instructions):

## CodeFire MCP

CodeFire provides persistent memory across Claude Code sessions via MCP.
Tasks, notes, and project context survive between conversations.

### Session Workflow

**Every session MUST start with:**
1. `get_current_project` — confirm which project is auto-detected
2. `list_tasks(status: "in_progress")` — resume unfinished work
3. `list_tasks(status: "todo")` — know what's queued
4. `list_notes(pinned_only: true)` — read critical context
5. Brief the user on what you found

**Every session MUST end with:**
1. Update all task statuses accurately
2. `add_task_note` on anything unfinished
3. `create_note` session summary if significant work was done

### Task Management

| Tool | Purpose |
|------|---------|
| `list_tasks` | List tasks by status (todo/in_progress/done) |
| `create_task` | Create a new task (title required) |
| `update_task` | Update status, priority, title, description, labels |
| `add_task_note` | Add progress notes to a task |
| `get_task` | Get full task details + notes |

### Notes — Institutional Memory

| Tool | Purpose |
|------|---------|
| `list_notes` | List project notes (pinned_only: true for critical) |
| `create_note` | Create a persistent note (title + markdown) |
| `update_note` | Update note content |
| `search_notes` | Full-text search across notes |

### Context Search

`context_search` — hybrid vector + keyword search across the indexed codebase.
Types filter: function, class, block, doc, commit

### Browser Automation

| Tool | Purpose |
|------|---------|
| `browser_navigate` | Open a URL |
| `browser_screenshot` | Capture current page |
| `browser_click` / `browser_type` | Interact with elements |
| `browser_console_logs` | Capture JS console output |

### Git Operations

| Tool | Purpose |
|------|---------|
| `git_status` | Check working tree state |
| `git_diff` | View changes |
| `git_log` | View commit history |
| `git_stage` | Stage files |
| `git_commit` | Create a commit |
Codex ~/.codex/AGENTS.md

Add the following to your ~/.codex/AGENTS.md file (global instructions):

## CodeFire MCP

CodeFire provides persistent memory across coding sessions via MCP.
Tasks, notes, and project context survive between conversations.

### Session Workflow

**Every session MUST start with:**
1. `get_current_project` — confirm which project is auto-detected
2. `list_tasks(status: "in_progress")` — resume unfinished work
3. `list_tasks(status: "todo")` — know what's queued
4. `list_notes(pinned_only: true)` — read critical context
5. Brief the user on what you found

**Every session MUST end with:**
1. Update all task statuses accurately
2. `add_task_note` on anything unfinished
3. `create_note` session summary if significant work was done

### Tools Reference

Tasks: list_tasks, create_task, update_task, add_task_note, get_task
Notes: list_notes, create_note, update_note, search_notes
Search: context_search (hybrid vector + keyword, types: function/class/block/doc/commit)
Browser: browser_navigate, browser_screenshot, browser_click, browser_type, browser_console_logs
Git: git_status, git_diff, git_log, git_stage, git_commit
Gemini ~/.gemini/GEMINI.md

Add the following to your ~/.gemini/GEMINI.md file (global instructions):

## CodeFire MCP

CodeFire provides persistent memory across coding sessions via MCP.
Tasks, notes, and project context survive between conversations.

### Session Workflow

**Every session MUST start with:**
1. `get_current_project` — confirm which project is auto-detected
2. `list_tasks(status: "in_progress")` — resume unfinished work
3. `list_tasks(status: "todo")` — know what's queued
4. `list_notes(pinned_only: true)` — read critical context
5. Brief the user on what you found

**Every session MUST end with:**
1. Update all task statuses accurately
2. `add_task_note` on anything unfinished
3. `create_note` session summary if significant work was done

### Tools Reference

Tasks: list_tasks, create_task, update_task, add_task_note, get_task
Notes: list_notes, create_note, update_note, search_notes
Search: context_search (hybrid vector + keyword, types: function/class/block/doc/commit)
Browser: browser_navigate, browser_screenshot, browser_click, browser_type, browser_console_logs
Git: git_status, git_diff, git_log, git_stage, git_commit
CodeFire Agent AGENTS.md (project root)

Add the following to your project's AGENTS.md:

## CodeFire MCP

CodeFire provides persistent memory across coding sessions via MCP.
Tasks, notes, and project context survive between conversations.

### Session Workflow

**Every session MUST start with:**
1. `get_current_project` — confirm which project is auto-detected
2. `list_tasks(status: "in_progress")` — resume unfinished work
3. `list_tasks(status: "todo")` — know what's queued
4. `list_notes(pinned_only: true)` — read critical context
5. Brief the user on what you found

**Every session MUST end with:**
1. Update all task statuses accurately
2. `add_task_note` on anything unfinished
3. `create_note` session summary if significant work was done

### Tools Reference

Tasks: list_tasks, create_task, update_task, add_task_note, get_task
Notes: list_notes, create_note, update_note, search_notes
Search: context_search (hybrid vector + keyword, types: function/class/block/doc/commit)
Browser: browser_navigate, browser_screenshot, browser_click, browser_type, browser_console_logs
Git: git_status, git_diff, git_log, git_stage, git_commit
6

Add a project

CodeFire auto-discovers projects or you can add them manually.

Automatic discovery

When you use Claude Code in a project directory, CodeFire automatically detects it by scanning ~/.claude/projects/. Your project will appear in the sidebar within a few seconds.

Manual add

Click Add Project in the project sidebar and select any project directory.

7

Gmail integration (optional)

Sync emails into CodeFire and convert them to tasks.

Gmail integration requires a Google OAuth Client ID and Client Secret. You'll need to create these in the Google Cloud Console.

  1. 1. Go to Google Cloud Console > APIs & Services > Credentials.
  2. 2. Create a new OAuth 2.0 Client ID (application type: Desktop app).
  3. 3. Enable the Gmail API for your project in the API Library.
  4. 4. Copy the Client ID and Client Secret.
  5. 5. In CodeFire Settings, go to the Gmail tab and paste both values.
  6. 6. Click Add Gmail Account and authorize access in the browser window that opens.
  7. 7. Toggle Enable Gmail sync on. Adjust the sync interval as needed (default: every 5 minutes).

Whitelist rules: Set up whitelist rules to automatically turn emails from specific senders into tasks with priority levels. Go to the Gmail settings tab to configure patterns.

8

CodeFire Capture (optional)

Mirror your agent chat to your phone and capture tasks on the go.

  1. 1. Install the CodeFire Capture mobile companion app on your phone.
  2. 2. In CodeFire Settings, enable Allow remote control from mobile and pair your device.
  3. 3. Watch runs stream live, approve agent actions, and send captured tasks straight into your staging inbox.

Remote control is opt-in, per device, and revocable at any time. The desktop app stays the source of truth.

You're all set.

Open a project, pick an agent, and start working. Your agent now has persistent memory, task tracking, browser automation, and semantic code search — all through MCP.