Foreman
github.com/mastaan66/foreman- Category
- AI Agents
- Rank
- No. 2195Tools index
Previous survey · No. 2093 ·
- Pricing
- Open Source
- Type
- TOOL
- Use case
- Coding
- GitHub
- 3 stars
- Latest release
- v0.4.0
- Date
About
Foreman orchestrates a tiered team of AI coding agents on top of opencode. A director writes tickets, and a lead, coders, testers, drones and a librarian carry them out under hard budgets and gates. This limits spend and keeps work moving through defined review steps.
At a glance
Foreman sends coding tickets to different tiers of OpenCode agents, records their work and runs the ticket’s verification commands. It is worth trying if you have repeatable jobs and want to reserve a more expensive model for planning and review.
Good fit
Teams with separate model tiers and coding tasks that have concrete acceptance checks.
Consider something else when
A single agent and a test command may be enough for one-off edits.
Start here
In a disposable repository, use foreman doctor to check the setup. Create a ticket that adds one test, then use foreman run with --verify. Compare the worker report with the actual verification output and review the recorded spend.
Cost and requirements
The MIT-licensed CLI requires Node 22 or later and OpenCode. Model usage depends on the providers you configure.
Limits to know
Review the ticket’s commands before running it. Budget stops can leave partial edits, and a passing verification command covers only what it checks. Run the first trial without private code or production credentials.
Sources checked 2026-11-05. Source-based assessment.
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.