Tripl

Keep your product analytics honest.

Website · Docs · Live demo · Source

tripl is an open-source tracking plan that checks itself against your warehouse.

You describe the events your product is supposed to send. tripl reads the events that actually land in your warehouse and tells you when the two drift apart: an event nobody documented, an event that stopped arriving, a property whose values changed, or a release that broke the volume.

• One plan for every event, property and value. Changes go through branches and review.
• Built for AI agents too. An MCP server gives agents the plan, so they build dashboards on the right events instead of guessing.
• Anomalies with a cause. Seasonal baselines, schema and value drift, and release regressions, each broken down by the slice that moved.
• Alerts where your team works. Slack, Microsoft Teams, Telegram, PagerDuty, email, webhooks, Jira and Linear.
• In CI. tripl check and typed code generation from the plan.

No SDK, and your data stays with you. tripl is self-hosted and reads the warehouse you already have, read-only: ClickHouse, PostgreSQL, BigQuery, Databricks and Greenplum, with Snowflake, Amazon Redshift, Trino and Amazon Athena in preview.

• 👀 Look around: demo.tripl.io gives you a demo project of your own in one click.
• 🚀 Run it: follow the quick start.
• 🛠 Operator CLI: pip install tripl
• 🤖 MCP server: pip install tripl-mcp, see connecting an AI agent.

Issues, ideas and pull requests are welcome in tripl-io/tripl. Please read the contributing guide first. Report security issues privately, as described in the security policy.

_{The server and web app are licensed under AGPL-3.0-or-later. The CLI and MCP server are licensed under Apache-2.0.}