title: Use Self-Hosted CodeRabbit With Bitbucket Server

sidebar_label: Bitbucket Server

description: Instructions to self-host CodeRabbit and integrate it with Bitbucket Server.

sidebar_position: 4

The self-hosted option is only available for CodeRabbit Enterprise customers with 500 user seats or more. Please contact [CodeRabbit Sales](mailto:sales@coderabbit.ai) to learn more about the CodeRabbit Enterprise plan.

- **Username**: Set the username to "CodeRabbit" for easier identification (optional).

- **Profile Image**: Use the CodeRabbitAI logo for the user image (optional).

Add the CodeRabbit user to each project where you want CodeRabbit to post reviews, with permissions to:

- Post reviews

- Open pull requests

Generate a personal access token for the CodeRabbit user to be added in the `.env` file as `BITBUCKET_SERVER_BOT_TOKEN`.

1. **Navigate to Webhook Settings**: Go to the repository settings and locate the webhooks configuration page.

2. **Configure Events**: Enable the following Pull Request events:

- "Opened"

- "Modified"

- "Comment Added"

3. **Add Webhook URL**: Enter the URL pointing to the CodeRabbit service, followed by `/bitbucket_server_webhooks` (e.g., `http://127.0.0.1:8080/bitbucket_server_webhooks`).

Create an `.env` file with the following content:

LLM_PROVIDER=openai

LLM_TIMEOUT=360000

OPENAI_API_KEYS=<openai-key>

OPENAI_BASE_URL=[<openai-base-url>]

OPENAI_ORG_ID=[<openai-org-id>]

OPENAI_PROJECT_ID=[<openai-project-id>]

LLM_PROVIDER=azure-openai

LLM_TIMEOUT=360000

AZURE_OPENAI_ENDPOINT=<azure-openai-endpoint>

AZURE_OPENAI_API_KEY=<key>

AZURE_GPT4OMINI_DEPLOYMENT_NAME=<gpt-4o-mini-deployment-name>

AZURE_GPT4O_DEPLOYMENT_NAME=<gpt-4o-deployment-name> # modelVersion: 2024-08-06

AZURE_O1_MINI_DEPLOYMENT_NAME=[<o1-mini-deployment-name>] # Optional

AZURE_O1_DEPLOYMENT_NAME=[<o1-preview-deployment-name>] # Optional: Provides best reviews but very expensive

AZURE_GPT4TURBO_DEPLOYMENT_NAME=[<gpt-4-turbo-deployment-name>] # Optional: modelVersion: turbo-2024-04-09

TEMP_PATH=/cache

AST_GREP_RULES_PATH=/home/jailuser/ast-grep-rules

AST_GREP_ESSENTIALS=ast-grep-essentials

SELF_HOSTED=bitbucket-server

BITBUCKET_SERVER_URL=<bitbucket-server-url>/rest

BITBUCKET_SERVER_WEBHOOK_SECRET=<webhook-secret>

BITBUCKET_SERVER_BOT_TOKEN=<personal-access-token>

BITBUCKET_SERVER_BOT_USERNAME=<bot-user-username>

CODERABBIT_LICENSE_KEY=<license-key>

CODERABBIT_API_KEY=[<coderabbitai-api-key>]

ENABLE_LEARNINGS=[true]

ENABLE_METRICS=[true]

JIRA_HOST=[<jira-host-url>]

JIRA_PAT=[<jira-personal-access-token>]

LINEAR_PAT=[<linear-personal-access-token>]

- If you are using Azure OpenAI, verify that the model deployment names are in the .env file.

- Values marked with [] are not optional to provide.

- You can generate `CODERABBIT_API_KEY` from CodeRabbit UI -> Organizations Settings -> API Keys.

Authenticate and pull the Docker image using the provided credentials file:

cat coderabbit.json | docker login -u _json_key --password-stdin us-docker.pkg.dev

docker pull us-docker.pkg.dev/coderabbitprod/self-hosted/coderabbit-agent:latest

You can query `/health` endpoint to verify that the coderabbit-agent service is up and running.

curl 127.0.0.1:8080/health

You can host the image on a server, serverless function, or container environment and expose port `8080`. Run the Docker image with the equivalent command on your chosen platform, ensuring you replace the `.env` file path with the path to your actual `.env` file:

docker run --env-file .env --publish 127.0.0.1:8080:8080 us-docker.pkg.dev/coderabbitprod/self-hosted/coderabbit-agent:latest