Hi HN, Yao here. Most frameworks are built for one developer, one agent. The moment a team shares infrastructure, people end up duct-taping row-level access into the prompt and hoping the model doesn't argue past it. I got tired of that.
Cast is a harness for multi-user, multi-agent systems: one server, a handful of people with their own identities, a fleet of agents handling different things and talking to each other when they need to. Agents are skills and CLAUDE.md, not Python classes, so you can focus on launching quick and refining the agent based on real usage. MIT, self-hosted, runs on a Mac Mini.
Cast puts access control in the routing layer, not the prompt. Each agent runs in its own container with actual filesystem boundaries. Identity verified before the agent sees the conversation (Slack, telegram, etc). Credentials never mounted in.
Developer alpha. Looking for teams that have hit the multi-user Claude Code wall and want to try this out. github.com/yaodub/cast. MIT. BYO Claude key.
embedding-shape 25 minutes ago [-]
> people end up duct-taping row-level access into the prompt
What exactly do you mean with this? The times I've collaborated on projects where most of us are using agents, we basically placed shared files in shared repositories, just like you usually do, so any shared instructions would go there. Then you work on your thing, then eventually submit a PR, and so on. Where does the "duct-taping row-level access" come into play, and how does it relate to the prompts themselves?
> MIT, self-hosted, runs on a Mac Mini.
Interesting approach to write something specifically for macOS and specifically for a Mac Mini :) I'm assuming this actually runs on whatever that can run JavaScript, right? :)
zwigglers 13 minutes ago [-]
For a dev team using agents as coding tools + coordinating via git, that workflow makes sense.
I built cast for other (non-coding) scenarios. A shared agent that multiple people interact with conversationally in real time, with different permission levels.
Think a household assistant on Telegram, or a small team's internal tool where sales and engineering collaborate but shouldn't see each other's data. There's no PR workflow there, just people chatting with a shared service.
On Mac Mini: Runs on anything with Node and a container runtime. Just trying to tap into the zeigeist.
binarymax 19 minutes ago [-]
The repo is showing initial commit of the entire codebase "6 minutes ago", and this was posted an hour ago? Are we time travelling now?
zwigglers 9 minutes ago [-]
You caught me. no time travel. I updated the README about an hour in (added the before/after code block). Codebase was there from the start, just some docs polish.
raphaelschaad 1 hours ago [-]
Cool — multi-user, multi-agent, virtual envs, single machine, is probably where most agentic DIY systems need to be going!
zwigglers 20 minutes ago [-]
[dead]
gmmeyer 2 hours ago [-]
this looks very cool, it’s genuinely hard to manage all my claudes sometimes and i’ve tried to build something like this as well, though with much less success
zwigglers 2 hours ago [-]
Thanks, means a lot. the "tried to build it myself" crowd is exactly who this is for
Rendered at 15:14:01 GMT+0000 (Coordinated Universal Time) with Vercel.
Cast is a harness for multi-user, multi-agent systems: one server, a handful of people with their own identities, a fleet of agents handling different things and talking to each other when they need to. Agents are skills and CLAUDE.md, not Python classes, so you can focus on launching quick and refining the agent based on real usage. MIT, self-hosted, runs on a Mac Mini.
Cast puts access control in the routing layer, not the prompt. Each agent runs in its own container with actual filesystem boundaries. Identity verified before the agent sees the conversation (Slack, telegram, etc). Credentials never mounted in.
Developer alpha. Looking for teams that have hit the multi-user Claude Code wall and want to try this out. github.com/yaodub/cast. MIT. BYO Claude key.
What exactly do you mean with this? The times I've collaborated on projects where most of us are using agents, we basically placed shared files in shared repositories, just like you usually do, so any shared instructions would go there. Then you work on your thing, then eventually submit a PR, and so on. Where does the "duct-taping row-level access" come into play, and how does it relate to the prompts themselves?
> MIT, self-hosted, runs on a Mac Mini.
Interesting approach to write something specifically for macOS and specifically for a Mac Mini :) I'm assuming this actually runs on whatever that can run JavaScript, right? :)
I built cast for other (non-coding) scenarios. A shared agent that multiple people interact with conversationally in real time, with different permission levels.
Think a household assistant on Telegram, or a small team's internal tool where sales and engineering collaborate but shouldn't see each other's data. There's no PR workflow there, just people chatting with a shared service.
On Mac Mini: Runs on anything with Node and a container runtime. Just trying to tap into the zeigeist.