Columna — an open-source data framework

skip to content Columna<br>An open-source data framework — a data model, a query language, and an engine — built so that no layer of it can return a confident wrong number.<br>pip install columna quickstart, two minutes →<br>start here<br>Never Let Your Agent Touch the Database Why your agent should never hold database credentials — and what to put between them. Three minutes.

transactionbasis: eventsinventorybasis: spinecategory_profilebasis: spine=day · storeproduct ↔ categorytouchprimarysplit<br>the demo manifold, drawn from the released package at build time.

Columna has three parts. The Manifold is a data model that holds what your data means — declared by a person, then checked against the data itself before it serves. FrameQL is a query language that can only ask for meaning the Manifold declares. The engine serves nothing the model can't defend.<br>Every answer comes back in one of four moods: serve (the number), disclose (the number and the assumptions it rides on), clarify (the question has more than one legitimate answer — choose), or refuse (the data does not define this — here's why). The same contract reaches a Python caller, an HTTP client, and an AI agent over MCP.<br>It is Apache-2.0, developed in the open, and currently at 0.13.2.

latest<br>release 0.13.2 · the declared Python floor and ceiling: 3.10–3.13, 64-bit Jul 27<br>position Never Let Your Agent Touch the Database Jul 25<br>paper The Two Great Sources of Silent Analytical Failure Jul 25<br>ledger The public roadmap and its open forks Jul 25

Try a question<br>The demo below runs the shipped package. Tap a question to see the answer and the mood it comes back in.

this is the actual wire — what an AI agent receives. Every answer is a pair: a result, and its disclosure. The mood is how the pair reads.<br>what's in this manifold? — explore the data before you ask<br>Browse the whole Manifold → the Explorer

👇 Tap a question to run it live. You'll get back the real result and the mood it answers in — served, disclosed, clarified, or refused.<br>the average of aov, by month — averaged from which grain? stock summed into months, per store last inventory, by customer — but inventory has no customers average order value (aov), by month

try to fool it — ask for a measure by name ask<br>this exact refusal, captured live, ships in the package.

live demo temporarily offline — this transcript was generated from the shipped package at build time (2026-07-28).

✦ New to the wire? Ask your own AI to explain it. Columna's "wire" is the answer format an AI agent receives from a data query: a RESULT paired with its<br>CONTEXT, in one of four moods — serve, disclose, clarify, refuse — on one contract (contract_version<br>"1"), identical over MCP or a direct API call. Example: asking `avg(aov) @ cal.month` — an average with<br>no pinned input grain — returns `clarify` (which grain to average from), never a guessed number. Explain<br>this wire format to me, and answer my follow-up questions about it. Copy prompt

Transcripts on this site are generated at build time by running the released package; they are not edited by hand.

Where to go<br>In the order a reader usually wants them.<br>Never Let Your Agent Touch the Database start here the position: why the model should never hold your database credentials, and what to put between them.

1 Run the demo install, and a clarify, a refuse, and a disclose print on your machine as wire JSON.

2 Read the case a complete worked example: a realistic warehouse, the Manifold that serves it, and recorded agent transcripts.

3 Learn the framework what Columna, FrameQL, and a Manifold are; the three manuals; the live Explorer.

4 Read the argument the eight-minute case for why this exists, with two live exhibits.

5 Positions and research every stance we hold with its evidence; six papers, the Silent Failure Atlas, and the public benchmark kit.

datumwise builds and stewards Columna: Huayin Wang (research), Irena Wang (engineering). It is an independent open-source research project.<br>We are taking three design partners for real-world deployments — contact@datumwise.ai.

pip install columna run it — two minutes