basin Skill (Cloudflare) for Claude: Stream Events Into R2 Tables and Query Them

Helps Claude build and fix Cloudflare Basin workflows: stream events into R2 Iceberg tables, manage the catalog and query it with SQL.

Installs: 25.8K; on skills.sh
Stars: 3.2K; on GitHub
Files: 16; besides SKILL.md
Updated: Oct 2026; last repo commit

This skill gives Claude a working map of Cloudflare Basin. Basin ingests and transforms events, stores them as Apache Iceberg tables in R2, and queries them with SQL. It covers three parts: Basin Pipelines to receive and transform events, Basin Catalog to manage the tables, and Basin SQL to query them.

SKILL.md says Cloudflare Data Platform is now Basin, and that Pipelines, R2 Data Catalog and R2 SQL are now Basin Pipelines, Basin Catalog and Basin SQL. Old resources and wrangler pipelines, wrangler r2 bucket catalog and wrangler r2 sql keep working. The skill points Claude to live Cloudflare docs for limits, pricing and syntax, because the new pages were still in an open pull request.

Good for

Developers sending event or log data into R2 and wanting it as queryable tables
Data engineers managing Iceberg catalogs and table maintenance on R2
Teams using the old names Pipelines, R2 Data Catalog or R2 SQL who need current guidance

Install basin

Claude Code

Copy this prompt.

Install the basin skill in Claude Code. Run this command:

npx skills add https://github.com/cloudflare/skills --skill basin

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

From its listing on skills.sh. Source.

What's in the skill

When it loads

Build and troubleshoot Cloudflare Basin analytics workflows with Basin Pipelines, Basin Catalog, and Basin SQL. Use for streaming data into R2 Iceberg tables, managing catalogs, or querying those tables; also use for requests using the former Data Platform, Pipelines, R2 Data Catalog, or R2 SQL names.

Sections in SKILL.md

Rebrand
Choose the workflow
Skill feedback

Key facts

Author: Cloudflare (the cloudflare/skills repo)
License: Apache-2.0 (as reported by GitHub)
What installs: 17 files including SKILL.md: 16 reference files: 5 each for pipelines, catalog and sql, plus skill-feedback.md; no scripts
Needs: Wrangler (npx wrangler basin ...), a Cloudflare account with R2, and WRANGLER_BASIN_SQL_AUTH_TOKEN for SQL
Docs status: SKILL.md says Basin docs and commands were still moving; check the live docs and your Wrangler version
Installs on skills.sh: 25,765 (Nov 2, 2026)
GitHub stars: 3,244 (Nov 2, 2026)
Last commit: 2026-10-31
skills.sh audits: Gen Agent Trust Hub pass, Socket pass, Snyk pass

What it includes

Receives events, transforms them with SQL and writes them to R2
A stream takes events by HTTP, a Worker binding or Logpush. A pipeline is SQL between the stream and a sink. A sink writes Iceberg tables via Basin Catalog or raw Parquet or JSON. `npx wrangler basin pipelines setup` creates all three.
Runs code · Basin Pipelines
Manages Iceberg tables, compaction, snapshot expiration and engine access
Enable it on a bucket with `wrangler basin catalog enable`. The guide separates the Iceberg REST catalog for query engines from Cloudflare's control-plane API.
Runs code · Basin Catalog
Read-only SQL over Iceberg tables in Basin Catalog
Run with `npx wrangler basin sql query` or the REST endpoint. There is no Workers binding for it. For writes the reference points to PySpark or PyIceberg.
Try: “Query the last 10 rows of my events table in Basin SQL”
Runs code · Basin SQL
Treats old names like R2 SQL and Pipelines as the new Basin products
Lets Claude answer questions that use the former names and give new commands for new work.
Rebrand mapping
Routes you to the pipelines, catalog or sql reference
Typical flow: Basin Pipelines, then Basin Catalog tables in R2, then Basin SQL or an outside engine.
Choose the workflow
Warns that schemas are fixed after creation and bad events are dropped quietly
The pipelines reference says stream schema, pipeline SQL and sink config cannot change after creation, and that first data takes 3 to 7 minutes to appear.
Gotchas and patterns
Can file a public GitHub issue about the skill after you approve the text
If you report wrong or outdated guidance, it drafts an issue for cloudflare/skills, shows you the exact title and body, and posts with gh issue create only after a yes.
Runs code · Skill feedback

From the official docs

Permissions

Medium risk

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

Deleting a stream removes buffered events
The pipelines API reference says deleting a stream removes buffered events and dependent pipelines. It also says `wrangler basin pipelines delete` answers "no" in non-interactive runs, so automation uses the REST DELETE. Check what depends on a stream first. github.com
Resources cannot be edited after creation
Stream schema, pipeline SQL and sink config are fixed, so a change means delete and recreate. Test in a scratch bucket. github.com
Feedback can become a public GitHub issue
It posts only after you see and approve the exact title and body. Leave account IDs, tokens and customer data out of it. github.com
Bad events are dropped without an error
Schema validation is deferred, so invalid events are accepted and then dropped. Watch the error metrics. github.com

Technical details

Talks to	Cloudflare APIs and developers.cloudflare.com docs. api.sql.cloudflarestorage.com for SQL queries. github.com, only for approved skill feedback
Sign-in and keys	Wrangler needs WRANGLER_BASIN_SQL_AUTH_TOKEN for SQL and does not reuse the wrangler login session. Store tokens as secrets.
Scopes	wrangler basin pipelines, catalog and sql commands against your Cloudflare account R2 API token with Admin Read and Write plus Basin SQL Read for SQL (per the SQL reference) gh issue create on cloudflare/skills, only after you approve the text

Questions

What does the basin skill do?

It helps Claude set up and troubleshoot Cloudflare Basin: sending events into R2 Iceberg tables, managing catalogs, and querying them with SQL. It also answers questions that use the old names.

Is Basin the same as R2 SQL and Pipelines?

SKILL.md says yes. Cloudflare Data Platform is now Basin, and Pipelines, R2 Data Catalog and R2 SQL are now Basin Pipelines, Basin Catalog and Basin SQL. Existing resources and the old Wrangler commands keep working.

How do I install the basin skill?

Use the install box above. Then you need Wrangler and a Cloudflare account with R2 to run the commands it suggests.

Does it need a Cloudflare API token?

For SQL, yes. The SQL reference says to set WRANGLER_BASIN_SQL_AUTH_TOKEN to an R2 token with Admin Read and Write plus Basin SQL Read, because Wrangler does not reuse its login session for Basin SQL.

Can Basin SQL write data?

No. The reference calls Basin SQL read-only and sends writes to PySpark or PyIceberg. Use it for analytics, not for transactions that need responses under 100 ms.

Why is my data not showing up?

The reference says the first flush takes 3 to 7 minutes. Invalid events are accepted and silently dropped, so also check the error metrics.

Is it safe?

It is text guidance with no scripts, but it leads Claude to run Wrangler commands that can create and delete resources and to handle API tokens. Review each command, and use a token limited to one bucket.

Does it send feedback to Cloudflare automatically?

No. If you report a problem with the skill, it drafts a public GitHub issue and posts only after you approve the exact text.

Information

updated Nov 2, 2026

Repository: cloudflare/skills
Folder: skills/basin

cloudflare/skills skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 6

github.com · github.com
basin on skills.sh · skills.sh
github.com · github.com
github.com · github.com
github.com · github.com
github.com · github.com