Skip to main content
Coda is a code companion that lives in Slack and works alongside your team. Most engineering work isn’t writing code. It’s understanding how things work, reviewing PRs, figuring out what broke, triaging the backlog, and deciding what’s safe to change. That work happens in Slack and GitHub, not in an editor. Coda fills that gap.

How It Works

Coda is a team of five specialist agents coordinated by a leader: Code isolation: All code changes happen in git worktrees on coda/* branches. Your main branch is never touched. Coda opens PRs; humans merge them.

Self-Learning

Coda improves without retraining through Agno’s LearningMachine. It picks up your coding standards, conventions, and patterns: Over time Coda stops waiting to be asked: it reviews open issues on a schedule, flags low-hanging fruit, and proposes changes.

Run Locally

See Setup guide for detailed instructions.
Confirm Coda is running at http://localhost:8000/docs.

Configure GitHub Access

Create a GitHub Personal Access Token with repo access and add it to .env:
Then tell Coda which repos to learn by editing repos.yaml:

Connect to the Control Plane

  1. Open os.agno.com
  2. Click Connect OSLocal
  3. Enter http://localhost:8000

Connect to Slack

Coda is designed to live in Slack as your teammate. Add your Slack credentials to .env:
Then restart to pick up the credentials:
See Slack setup for creating the Slack app.

Deploy to Railway

See Deploy to Railway for JWT setup and production configuration.
The script provisions PostgreSQL, configures environment variables, and deploys Coda. Then connect via the control plane:
  1. Open os.agno.com
  2. Click Connect OSLive
  3. Enter your Railway domain

Example Prompts

Try these after connecting your repos:
  • How does authentication work in this codebase?
  • Review the open PR on feature/new-api
  • What are the most important open issues?
  • Where is the database connection configured?
  • What testing conventions does this project use?

Source

For GitHub access setup, Slack configuration, and architecture details, see the GitHub repo.