GeezerVison

The World Wide Web, restored to 800 pixels of glory.

GeezerVison is showing rawquery.dev — Your data. In Claude Code. | rawquery

Managed data. Command-line control.

Your data.
In Claude Code.

Connect your sources. Ask Claude Code to explore your data, write SQL, and run it through rq. rawquery handles syncing, storage, and query execution.

Start free See the workflow ↓

$curl -fsSL rawquery.dev/install.sh | sh

Then rq login, rq claude-init, claude.

No credit card·Free plan included

Also works with Codex and Claude in your browser. Setup guide →

Recorded on the product · 2:00

0:00 / 2:00

Intro

rawquery from the rq CLI and Claude Code: connect an API, explore and query it with SQL, ask Claude Code a question about the data, publish a page.

Built-in connectors

+ any REST / GraphQL API via JSON spec

The agent works. You see the SQL.

Ask. Inspect. Run. Iterate.

Connect your first source →

Example prompt / Stripe invoices already synced

> Count paid invoices by month. Show the six most recent months, newest first. Check the schema before writing the query.

Inspect the source

rq describe stripe_live.invoices

Execute with rq query

SELECT strftime(created, '%Y-%m') AS month,
       COUNT(*) AS invoice_count
FROM stripe_live.invoices
WHERE paid = true
GROUP BY 1
ORDER BY 1 DESC
LIMIT 6

Recorded rq output / 2026-09-15

Paid invoice counts by creation month, returned by the displayed SQL.
month	invoice_count
2026-08	27
2026-07	63
2026-06	66
2026-05	56
2026-04	59
2026-03	266

Actual query results. Counts include zero-value paid invoices. Use your connection name in place of stripe_live.

> Now save the query and make it a chart.

See the published full history ↗

Claude Code discovers schemas and runs commands through the rq skill. Review the SQL, refine the question, keep going. Your agent uses your workspace permissions. Full walkthrough →

One CLI. The whole data workflow.

rq connections →

Bring your sources together.

Connect databases and SaaS tools. Schedule syncs. Query across sources with SQL.

rq transforms →

Keep the useful work.

Save queries, build SQL transforms, and schedule them. Pick up where you left off.

rq charts →

Ship the result.

Turn saved queries into charts and shareable pages. Your agent can build those too.

Run rq yourself, use it in scripts, or let your agent drive. Same commands. Explore the CLI →

Give your agent something to query.

Create a workspace. Connect a source. Open your terminal.

Start free →

Free plan. No credit card.Hosted in Frankfurt.DuckDB. Open Parquet storage.

You are visitor 01402503 — GeezerVison does not store the pages you look up.
Best viewed with Netscape Navigator 4.7 or Internet Explorer 6 at 800×600. No cookies required.