npm i -g @openai/codex
or brew install --cask codex

Codex CLI is a coding agent from OpenAI that runs locally on your computer.

If you want Codex in your code editor (VS Code, Cursor, Windsurf), install in your IDE.
If you are looking for the cloud-based agent from OpenAI, Codex Web, go to chatgpt.com/codex.

---

Quickstart

Installing and running Codex CLI

Install globally with your preferred package manager:

# Install using npm
npm install -g @openai/codex

# Install using Homebrew
brew install --cask codex

Then simply run codex to get started.

You can also go to the latest GitHub Release and download the appropriate binary for your platform.

Each GitHub Release contains many executables, but in practice, you likely want one of these:

• macOS
  • Apple Silicon/arm64: codex-aarch64-apple-darwin.tar.gz
  • x86_64 (older Mac hardware): codex-x86_64-apple-darwin.tar.gz
• Linux
  • x86_64: codex-x86_64-unknown-linux-musl.tar.gz
  • arm64: codex-aarch64-unknown-linux-musl.tar.gz

Each archive contains a single entry with the platform baked into the name (e.g., codex-x86_64-unknown-linux-musl), so you likely want to rename it to codex after extracting it.

Using Codex with your ChatGPT plan

Run codex and select Sign in with ChatGPT. We recommend signing into your ChatGPT account to use Codex as part of your Plus, Pro, Team, Edu, or Enterprise plan. Learn more about what's included in your ChatGPT plan.

You can also use Codex with an API key, but this requires additional setup.

Docs

• Codex Documentation
• Contributing
• Installing & building
• Open source fund

Next AI Labs Fork - Semantic Memory Integration

This fork adds production-ready semantic memory injection powered by Agent Swarm, enabling Codex to leverage your organization's knowledge base and past learnings. All features are opt-in and do not change upstream defaults.

🚀 What's Different?

This fork automatically injects relevant context from your Agent Swarm knowledge base into every prompt, enabling Codex to:

• Remember solutions from previous sessions
• Access organizational patterns and best practices
• Leverage team knowledge without manual copy-paste
• Learn from past successes and failures

Prerequisites:

• Rust toolchain (1.92.0+)
• Just command runner: cargo install just
• Agent Swarm MCP: git clone https://github.com/Next-AI-Labs-Inc/agent-swarm-mcp.git

Installation:

# 1. Clone this fork
git clone https://github.com/Next-AI-Labs-Inc/codex.git
cd codex

# 2. Point to your Agent Swarm installation
export AGENT_SWARM_PATH="/path/to/agent-swarm-mcp"

# 3. Verify Agent Swarm works
$AGENT_SWARM_PATH/scripts/swarm -v "test query"

# 4. Build and run Codex
just build
just codex

💡 How It Works

On each prompt, Codex:

1. Extracts key terms from your input
2. Runs semantic vector search: swarm -v --limit 5
3. Injects relevant memories if found (3s timeout)
4. Logs all injections to ~/.codex/log/memory_injection.log

Injected format:

Memories which may be helpful:
<relevant context from swarm>

Performance:

• Non-blocking with 3s timeout
• Only active when AGENT_SWARM_PATH is set
• Zero impact if disabled

Implementation:

• Core logic: codex-rs/core/src/memory_search.rs
• Integration: codex-rs/core/src/codex.rs:385-425

🧪 Testing

# 1. Test Agent Swarm directly
$AGENT_SWARM_PATH/scripts/swarm -v "authentication patterns"

# 2. Start Codex and ask a question
just codex
# Prompt: "How should I implement user authentication?"

# 3. Check injection log
tail ~/.codex/log/memory_injection.log

Example log output:

[2026-01-17 10:30:45 UTC]
Query: implement user authentication
Injected:
✓ Found 3 relevant memories about auth patterns

🛠️ Configuration

Disable memory injection:

unset AGENT_SWARM_PATH

Custom log location:

export CODEX_HOME="/custom/path"
# Logs to: $CODEX_HOME/log/memory_injection.log

Adjust memory limit (default: 5): Edit codex-rs/core/src/memory_search.rs:80:

.arg("--limit").arg("10")

🐛 Troubleshooting

Memory injection not working?

# 1. Verify environment variable
echo $AGENT_SWARM_PATH

# 2. Test swarm directly
$AGENT_SWARM_PATH/scripts/swarm -v "test"

# 3. Check logs
tail -f ~/.codex/log/memory_injection.log

📚 Upstream Compatibility

This fork tracks openai/codex:main and can be rebased:

git remote add upstream https://github.com/openai/codex.git
git fetch upstream
git rebase upstream/main

This repository is licensed under the Apache-2.0 License.