every day the @runpentagon desktop experience gets faster and more reliable
shoutout @s2_streamstore for helping our multiplayer agent-to-agent communication cook
turns out "one stream per session" is not just for agents.
it works for browser sessions too.
built RePlaya, a self-hosted session replay tool where each visit is just an @s2_streamstore stream you can play back or tail live.
recording = append
playback = read
live replay = tail
Streams as a cloud storage primitive, now in more places.
S2 is the serverless API for unlimited, durable, real-time streams. Today, we're adding US West and EU North as new basin locations.
https://t.co/rBU10H9wMr
@tonboio good to know of the catch-up breakage, we'll investigate and fix! s2-lite is much younger than the cloud service, it's the same API but different backends
@tonboio congrats on launching. multi-node s2-lite is not on the current roadmap :) for elastic scalability we recommend the managed service https://t.co/d5JGtMJpXI
Streaming shouldn’t break when connections do.
Meet Durable Global Streams, like Kafka and S3 combined into severless storage for streams: @s2_streamstore
→ Replayable
→ Resumable
→ Globally available, backed by Tigris storage
Built for real-time apps, AI agents, and anything that can’t afford to lose data.
Back S2 with Tigris 👇
S2 now supports client-supplied encryption keys for durable streams.
Bring your own per-request key, keep key material out of S2 storage and logs, and let S2 handle request-time encryption behind the stream API.
Claude makes for a great curl console. No MCP needed 😅
Here's a prompt to get you started with @s2_streamstore (sign up and get an access token first, see env var reference)
```
Refer to https://t.co/iZvTd2tAEF and use S2_ACCESS_TOKEN env as bearer token. Create a basin called something unique with stream auto-creation enabled on read or append, and delete empty streams after 1h. Start an SSE read and append a "hello world" record, let me know if it comes through.
```
@garrytan put a @runpentagon gstack team on this - here's a skill to set up private group chats for agents built on top of @s2_streamstore:
shoutout @sgbalogh0 for a great coffee chat the other day that put s2 top of mind
https://t.co/DJIYDJVI46
There's two new ways to manage S2 as code! The Terraform provider lets you manage basins, streams, and access tokens as resources with full lifecycle support. Available on the Terraform Registry at s2-streamstore/s2.
Or you can use the CLI using `s2 apply` which takes a declarative JSON spec file and creates or reconfigures the specified resources. You can also run it with `--dry-run` to preview a field-level diff before applying, and use `--schema` to get JSON Schema for editor autocomplete and validation.
You can also use either with s2-lite for local dev and CI!