Datris
Connect your agent to a Datris pipeline for ingest, validation, transforms, and queries without hand-writing every data-ops script.
Overview
io.github.datris/datris is a Build-phase MCP server that lets agents ingest, validate, transform, and query data through a configured Datris pipeline URL.
What is this MCP server?
- Agent-native data platform tools: ingest, validate, transform, and query via MCP
- PyPI datris-mcp-server 1.8.4 stdio plus optional Docker OCI and localhost streamable-http on port 3000
- Requires PIPELINE_URL to your Datris pipeline server; PIPELINE_API_KEY when key validation is on
- Published version 1.8.4 from datrisai/datris-mcp-server images and PyPI identifier
- Fits builders who centralize data movement in Datris rather than ad hoc notebooks
- MCP server version 1.8.4
- PyPI package identifier datris-mcp-server with stdio transport
- Optional streamable-http at http://localhost:3000/mcp
Community signal: 10 GitHub stars.
What problem does it solve?
Your agent cannot safely drive your data pipeline because ingest and validation steps live in a separate UI and undocumented APIs.
Who is it for?
Solo builders already committed to Datris pipelines who want agent-driven data workflows with stdio, Docker, or local HTTP options.
Skip if: Greenfield projects with no Datris deployment or teams that only need a single read-only SQL MCP.
What do I get? / Deliverables
After PIPELINE_URL and optional PIPELINE_API_KEY are set, MCP tools can orchestrate Datris pipeline operations from your coding agent session.
- MCP-accessible ingest, validation, transform, and query operations against Datris
- Repeatable stdio, Docker, or streamable-http MCP deployment patterns
- Agent-driven data workflow automation aligned with Datris 1.8.4 server
Recommended MCP Servers
Journey fit
Primary shelf is Build because most solo builders first encounter Datris while connecting a product or agent to a managed pipeline. Integrations reflects MCP as the bridge to an external Datris pipeline service rather than in-repo application code.
How it compares
Full pipeline orchestration MCP, not a lightweight CSV reader or warehouse metadata-only server.
Common Questions / FAQ
Who is io.github.datris/datris for?
Indie developers and small teams running Datris who want Claude Code, Cursor, or Codex to operate ingest, validation, transform, and query steps.
When should I use io.github.datris/datris?
Use it while building data-backed features and whenever you need agent-assisted pipeline changes tied to a live PIPELINE_URL.
How do I add io.github.datris/datris to my agent?
Install datris-mcp-server 1.8.4 via PyPI stdio, Docker OCI, or streamable-http, set PIPELINE_URL (and PIPELINE_API_KEY if required), then register the server in your MCP config.