Open SWE - An Open-Source Asynchronous Coding Agent

Open SWE is an open-source cloud-based asynchronous coding agent built with LangGraph. It autonomously understands codebases, plans solutions, and executes code changes across entire repositories—from initial planning to opening pull requests.

Note: you're required to set your own LLM API keys to use the demo.

Note

💬 Read the announcement blog post here

Features

• 🔗 Trigger from Linear, Slack, or GitHub — mention @openswe in a Linear comment, Slack thread, or GitHub PR comment to kick off a task
• 👀 Instant acknowledgement — reacts with 👀 the moment it picks up your message so you know it's on it
• 💬 Message it while it's running — send follow-up messages mid-task and it'll pick them up before its next step
• 🔀 Run multiple tasks in parallel — each task runs in its own isolated cloud sandbox, no queuing
• 🔐 GitHub OAuth built-in — authenticates with your GitHub account automatically, no token setup needed
• 🚀 Opens PRs automatically — commits changes and opens a draft PR when done, linked back to your Linear ticket

Installation

Prerequisites

• Python 3.11+
• uv package manager
• LangGraph CLI
• ngrok (for exposing local webhooks)

1. Clone the repo

git clone https://github.com/langchain-ai/open-swe.git
cd open-swe/apps/agent

2. Install dependencies

uv sync

3. Set up the Linear webhook

In a terminal, start ngrok to get your public URL:

ngrok http 8000
# e.g. https://xxxx.ngrok.io

Then in Linear:

1. Go to Settings → API → Webhooks → New webhook
2. Fill in:
   • Label: open-swe
   • URL: https://xxxx.ngrok.io/webhooks/linear
   • Secret: generate a random string — copy it, you'll need it for LINEAR_WEBHOOK_SECRET
3. Under Data change events, enable Comments → Create only
4. Click Create webhook

To get your LINEAR_API_KEY:

1. Go to Settings → API → Personal API keys → New API key
2. Name it open-swe and copy the key

4. Set environment variables

Create a .env file in apps/agent/ with the following:

# LangSmith
LANGSMITH_API_KEY_PROD=""           # Your LangSmith API key
LANGCHAIN_TRACING_V2="true"
LANGCHAIN_PROJECT=""

# LLM
ANTHROPIC_API_KEY=""                # Anthropic API key (recommended default provider)

# GitHub OAuth (via LangSmith agent auth)
GITHUB_OAUTH_PROVIDER_ID=""         # GitHub OAuth provider ID from LangSmith
X_SERVICE_AUTH_JWT_SECRET=""        # Secret for service JWT tokens

# GitHub App (Bot)
GITHUB_APP_ID=""                    # GitHub App ID
GITHUB_APP_PRIVATE_KEY="-----BEGIN RSA PRIVATE KEY-----
...
-----END RSA PRIVATE KEY-----
"
GITHUB_APP_INSTALLATION_ID=""       # GitHub App installation ID

# GitHub Webhook
GITHUB_WEBHOOK_SECRET=""            # Secret for verifying GitHub webhooks

# Linear
LINEAR_API_KEY=""                   # Linear API key (from step 3)
LINEAR_WEBHOOK_SECRET=""            # Secret you set when creating the webhook (from step 3)

# Slack (optional)
SLACK_BOT_TOKEN=""
SLACK_BOT_USER_ID=""
SLACK_BOT_USERNAME=""
SLACK_SIGNING_SECRET=""

# Sandbox
DEFAULT_SANDBOX_TEMPLATE_NAME=""    # LangSmith sandbox template name (uses default if not set)

# Token encryption
TOKEN_ENCRYPTION_KEY=""             # 32-byte url-safe base64 key for encrypting GitHub tokens

5. Run the agent

In one terminal, start the LangGraph dev server:

uv run langgraph dev --no-browser

In a second terminal, start the webhook server:

make run

The LangGraph server runs on http://localhost:2024 and the webhook server on http://localhost:8000.

6. Verify it works

Comment @openswe on any Linear issue. You should see:

• A 👀 reaction on your comment within a few seconds
• A new run appear in your LangSmith project

---

Usage

Open SWE can be used in multiple ways:

• 📋 From Linear. Mention @openswe in a comment on any Linear issue and describe the task you want it to perform (e.g. @openswe fix the login bug described above). The agent will pick up the issue context along with your instructions and start working on it.
• 🐙 GitHub (for Open SWE-generated PRs). Once Open SWE completes implementation, it pushes changes to a branch open-swe/<thread-id> and opens a draft pull request linking back to the originating Linear issue. From there, you can review the code, request changes, and merge when ready.