Archive

Run a fleet of coding agents while keeping branches, reviews, and CI failures manageable.

Every project gets a main agent. Tell it what you want built and it plans the work, starts a worker for each piece in its own worktree, and comes back only when something needs you.

Sessions place themselves from live facts: what the agent is doing, its pull request, CI, and reviews. You open what needs you and leave the rest alone.

Someone mentioned this plannotator inbox in response to Ryan Dahl's (node, and deno creator) idea that Emails are a better way to talk to coding agents than IRC like interface.

talking to a coding agent is more like email than irc - because the interactions are few and far between, I often finding myself quoting their most recent text, answering multiple questions. addressing different topics. it's like replying to an email.

It makes me wonder if you could have a coding agent UI that was more like an email client or mailing list kind rather than chat

I agree with this because email is a better medium for this:

1. asynchronous mode of communication
2. can include rich media in repose
3. can copy other workers (both bot, and humans) on the thread as needed
4. Access control, traceability, and other "audit" concerns are already handled well on email.
5. You can think of coding agent as another "remote" employee that you have seen outside of emails.

This is an idea worth pulling on.

Dory is the complete local development system for Apple Silicon: Docker, Compose, Kubernetes, virtual machines, and policy-bound agent sandboxes.

Terminal-native coding agent running on Kimi K2.6 via Cloudflare Workers AI. Bring your CloudFlare key, install the CLI, start coding. multi-turn tool use, 262k context, vision inputs, full agentic loop. one key, one bill. Cheaper per token than most hosted coding assistants and fully transparent about what it’s doing.