foreman

Name: foreman
Rating: 63 (1 reviews)
Author: VisionForge-OU

VisionForge-OU/foreman

Foreman is a Boris-style agentic orchestrator with a Textual terminal UI that supervises headless Claude Code agents through a gated software-delivery pipeline.

484 102No licenseUpdated 2026-06-29

Good

View on GitHub Open source file

Overview

Foreman acts as a supervisor for autonomous coding agents, managing software delivery from planning and ADR/PRD creation to issue tracking, TDD builds, and end-to-end verification. It features a keyboard-driven TUI, database-free human-readable file persistence inside target repositories, git worktree isolation, strict per-run token and cost budgets, and a built-in evaluation flywheel.

Capabilities

▸Gated software delivery pipeline (plan -> ADR/PRD -> issues -> TDD build -> e2e)
▸Headless Claude Code CLI execution with stream-json event parsing
▸Per-run turn, cost, and time budget enforcement
▸Keyboard-driven Textual terminal user interface (TUI)
▸Git worktree isolation for parallel worker execution
▸Crash-safe filesystem-based state persistence without external databases
▸Automated failure clustering and retro evaluations

Best for

Orchestrating multi-step software development features with AI agents, Supervising autonomous Claude Code instances with human review gates, Running isolated parallel development tasks in separate git worktrees, Managing structured software development workflows from plan to e2e test

Works with

Claude Code

Evoa Score breakdown= Σ (score × weight)

Task usefulness20%100 → +20.0

実タスクにどれだけ役立つか（機能の豊富さ・用途の明確さ）。 — AIによるcapabilities/use-cases解析

Code quality15%90 → +13.5

実装・指示の品質。 — AIによるSKILL.md/README解析

Maintenance15%85 → +12.8

リポジトリがどれだけ活発に保守されているか。 — GitHub 最終push日時の新しさ

Documentation12%40 → +4.8

ドキュメントの充実度・分かりやすさ。 — README/独自要約の情報量

Security15%50 → +7.5

危険・不審な挙動が無いか。 — AIによるセキュリティレビュー

Originality10%85 → +8.5

ありふれたラッパーではない独自性。 — AIによる独自性判定

Popularity8%71 → +5.7

コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）

Compatibility5%25 → +1.3

対応AIエージェントの広さ。 — AIによる対応エージェント判定

Weighted total74.0 / 100

ライセンス不明/制限あり（Red）のSkillは総合スコアに0.85倍の補正を適用します。ランキングはこのScoreのみで決まり、広告で変わりません。算出方法の詳細 →

Security considerations

Foreman invokes local headless Claude CLI agents and runs shell commands/tests during the TDD build phase. It enforces per-run cost limits, hard daily stops, and deny hooks to prevent unauthorized tool use, but users must exercise caution regarding target repository code execution.