We launched @amikadev: the open source operating system for AI agents to automate coding and prove the code correct
Amika runs fleets coding agents in cloud VMs to automate flows like:
- fix production errors, with a test proving the fix
- ship user feature requests from Slack
I run my GTM with markdown files and claude/codex, but I missed some of the structure you get from using a SQL database
So I made a open source CLI tool called "akfs" to work with the YAML frontmatter in markdown files
In UNIX tradition, it's composable
A software factory needs 5 things:
1. A workstation for every agent
2. Rules wired into hooks, not just AGENTS.md
3. Feedback that runs while the agent works
4. A workflow control plane
5. Agentic CI that fixes, not just passes/fails
@amikadev shipped a Slackbot that spawns coding agents and gives each its own sandboxed computer.
Runs any model + harness inside a Linux sandbox. It connects to your repos and can run and test your actual apps per sandbox.
We can configure different agents per channel. Bug-fix channel gets a bug-fix agent. Every Slack Connect channel from a customer gets a feature-request agent.
Any other good Slack automation ideas?
Built an agent to auto-implement Linear issues in parallel
It spins up sandboxed coding agents on @daytonaio, and makes a PR for each issue and watches Github Actions and auto-fixes any problems
Creating and running coding sandboxes should be as easy as `git clone` and `git push`.
We're designing so you can:
1. use it from your CLI
2. integrate it into tmux or an agent development environment (ADE)
3. call it as an API from you CI
Your entire SDLC should be automatable with AI agents. But just like there's no one-stop dev workflow, there's no one-stop agent.
That's why we're building https://t.co/3fuBXhgLe6: infra for every company to build their own software factory, customized to them.