Integrations
Each integration is a Pydantic AI Harness capability that gives an agent the tools of one outside service.
Most connect to the service’s own hosted MCP server, and the credential you pass decides what the agent can reach. Add
one to capabilities=[...] like any other capability.
- GitHub: read and change repositories, issues, pull requests, and other GitHub resources.
- Linear: read and change Linear issues, projects, teams, and comments.
- Macroscope: run a local Macroscope code review and hand the findings to the agent.
- LocalStack: provision and exercise AWS services in an emulated AWS environment, without a real account.
- Google Workspace: use Gmail, Calendar, Drive, and other Google Workspace products.
- Notion: search and change content in a Notion workspace.
- Slack: read and send Slack messages, browse channels, and work with canvases.
- Day AI: search and update CRM records, read meeting context, and draft emails in Day AI.
- Grain: search and read meetings, transcripts, notes, and deals in Grain.
- Pylon: work with support issues, accounts, and contacts in Pylon.
- StackOne: act on a user’s linked business applications, such as BambooHR, Salesforce, or Zendesk, through StackOne.
- PostHog: query product analytics and manage feature flags, experiments, and dashboards in PostHog.
- Logfire MCP: query Logfire telemetry and manage Logfire projects.
If the service runs an MCP server, connect it with the core MCP capability. For tools from other libraries, see Third-Party Tools.