foreman

# <img src="https://raw.githubusercontent.com/thruwire/foreman/main/docs/assets/foreman-icon.png" alt="" width="32"> Foreman

Foreman watches the software factory floor with decision models, placing a fast semantic supervisor
above slower coding agents. It supports TypeSafe AI's Jev and OpenAI's Decisions API.

Give it a ticket, specification, bug report, or any free-form software job. Foreman can run
[Codex](https://learn.chatgpt.com/docs/developer-commands?surface=cli),
[OpenCode](https://opencode.ai), [Hermes](docs/workers.md), or
[LangChain's Deep Agents Code (dcode)](integrations/deepagents/README.md) as workers.
It can also attach supervision to supported coding-agent sessions, including
[Pi and Pi Durable](integrations/pi/README.md).

The coding agent does the software engineering. Foreman independently assesses completion and
requirements against the original job. It also checks whether test evidence is sufficient and
whether verification or human input is needed. You can configure additional checks and
responsibilities for your workflow.

```text
                         SOFTWARE FACTORY
         Coding agent     Coding agent           Tests
            worker             worker                │
               │                  │                  │
               └──────────────────┼──────────────────┘
                                  │ factory evidence
                                  ▼
                              FOREMAN
                          decision model
                                  │
                                  ▼
                    responsibilities
                      completion
                        implementation_complete  .91
                        requirements_satisfied   .79
                        ready_to_finish          .