i'm hiring a pm for @cloudflare r2 if you know anyone who would be a good fit lmk

the biggest voice / video chat app you know runs on @cloudflare compute discord.com/blog/how-we-move…

Cloudflare continues to build up its zero-egress-charges data warehouse capabilities with R2 Pipelines Ingest data into an Data Catalog-enabled R2 bucket, using Pipelines or other mechanisms, then run queries over it using R2 SQL or your favourite engines like Snowflake. Pipelines can ingest, transform, and load streaming data into Apache Iceberg or Parquet in R2. Both got new features on Monday: You can now ingest logs directly into Pipelines from your Cloudflare Workers by using Pipelines as a Logpush destination. Logs can be noisy, so using Pipelines, you can trim down to the fields you need before the data is stored in R2 in Parquet files or Apache Iceberg tables. R2 SQL now supports functions for easily querying JSON data, alongside improving readability of EXPLAIN queries by allowing them to be formatted as JSON. Lastly, unpartitioned Iceberg tables can now be queried directly - this should only be used for smaller tables, as partitions will significantly help performance at larger scale.

Last week we shipped the R2 Data Catalog UI so now users have a single spot to create catalogs, see metrics, schema evolution diffs, metadata, and manage compaction/snapshot expiration settings from the Dash. Made possible with a new API route that gives you the metadata from your Iceberg table snapshots without having to go through a client like Spark. Makes it much easier for your agents to get all the good info about your data without having to go through extra steps. GET "api.cloudflare.com/client/v4…{account_id}/r2-catalog/{bucket_name}/namespaces/{namespace}/tables/{table_name}" ... Next stop: How about a way to query your data directly from the Dash?

serverless was a mistake

agent harnesses are like web frameworks let a thousand flowers bloom, no one-size-fits-all, right tool for the job. there are and will be many great ones what agent harnesses have in common — they all benefit from using the Agents SDK to run in Durable Objects, the runtime designed for agents npm i agents agents.cloudflare.com

Introducing Flue — The First Agent Harness Framework Flue is a TypeScript framework for building the next generation of agents, designed around a built-in agent harness. Flue is like Claude Code, but 100% headless and programmable. There's no baked in assumption like requiring a human operator to function. No TUI. No GUI. Just TypeScript. But using Flue feels like using Claude Code. The agents you build act autonomously to solve problems and complete tasks. They require very little code to run. Most of the "logic" lives in Markdown: skills and context and AGENTS.md. Flue is like Astro or Next.js for agents (not surprising, given my background 🙃). It's not another AI SDK. It's a proper runtime-agnostic framework. Write once, build, and deploy your agents anywhere (Node.js, Cloudflare, GitHub Actions, GitLab CI/CD, etc). We originally built Flue to power AI workflows inside of the Astro GitHub repo. But then @_bgiori got his hands on it, and we realized that every agent needs a framework like Flue, not just us. Check it out! It's early, but I'm curious to hear what people think. Are agents ready for their library -> framework moment?

Music-2.6 from @MiniMax_AI is free on Cloudflare this week! Generate full-length songs or instrumentals from a text prompt, with optional auto-generated lyrics. Fast global inference on Cloudflare’s network. Perfect for building music apps on Cloudflare Workers. Try it free before the promo ends: developers.cloudflare.com/ai…