Integrations
Integrations allow agents to interact with external tools and services, extending their capabilities beyond chat.
Available Integrations
| Integration | Agents | Use Case |
|---|---|---|
| GitHub | OpenClaw (planned) | PR reviews, issues |
| Jira | OpenClaw (planned) | Issue tracking |
| GitLab | Not planned | Alternative to GitHub |
| Confluence | Not planned | Documentation |
| Linear | Not planned | Issue tracking |
| Notion | Not planned | Knowledge base |
Integration vs Provider vs Channel
| Component | Purpose | Example |
|---|---|---|
| Provider | AI model backend | OpenAI GPT-4 |
| Channel | How users talk to agent | Discord, CLI |
| Integration | What the agent can do | Read GitHub issues |
How Integrations Work
Integrations typically:
- Read data: Fetch information from external services
- Take actions: Create/update items in external services
- Receive events: Webhooks for real-time updates
- Use credentials: API keys or OAuth tokens
Configuration
Integrations will be configured during agent onboarding or separately:
# Future command (not yet available)
clm agent configure <agent-name> --integrations
Security
- API tokens stored securely
- Scoped permissions (least privilege)
- Audit logging for actions
- Per-integration credential isolation
Requesting New Integrations
To request a new integration:
- Check if it exists in 📋 Not Planned section
- Open an issue describing:
- The service (e.g., PagerDuty, Zendesk)
- Use case
- Required actions (read, write, both)
- Consider contributing a PR
See individual integration pages for detailed setup instructions.