An AI coding agent controlled like a machine.
Auto Code Machine treats LLM code generation as a controlled process—not free-form writing. Inspired by mechanical control engineering, it uses design rules, small work orders, and checkpoints to keep output reliable, reviewable, and reversible.
- Design first — agree on scope and architecture before coding
- Try a small slice first — prove the design with a minimal end-to-end path
- Small work orders — split work into verifiable steps, generated as you go
- Gates by risk — light checks for safe changes, full review for risky ones
- Fix by feedback — let tests and linters guide repairs
- Remember failures — don't repeat the same mistake
Each work order is one small, verifiable task—not one tiny code edit. It defines the goal, owner, allowed files, how to verify, and how to roll back.
| Engineer | Tier | Role |
|---|---|---|
| Chief Engineer | Reasoning | Orchestration and risk calls |
| Design Engineer | Reasoning | Product and architecture baselines |
| Planning Engineer | Execution → Reasoning | Break work into orders |
| Code Engineer | Execution | Write code within scope |
| Test Engineer | Execution | Tests and acceptance criteria |
| Review Engineer | Reasoning | Review diffs, approve or block |
| Security Engineer | Reasoning | Auth, permissions, secrets |
| Repair Engineer | Execution | Fix issues from test feedback |
| Memory Engineer | Execution | Record failures and decisions |
- Low — docs, comments, trivial changes. Automated checks only.
- Normal — regular features. Checks + review.
- High — auth, payments, migrations. Checks + review + security.
- Exploration — debugging. Observe first, change little.
Tuned for cache-friendly models like DeepSeek and MiMo to keep cost down.
- Tier routing — fast models for coding, stronger models for decisions
- Stable context — structured prompts to improve cache hits
- Tool call repair — auto-fix broken model tool calls
- Cost tracking — spend per turn, order, and session
See CONTRIBUTING.md.