CodeMachine
Turn repeatable AI coding processes into long-running, coordinated workflows.
- Source repo
- moazbuilds/CodeMachine-CLI
- Stars
- ★ 2.5k
- Last updated
- 8mo ago
- License
- Apache-2.0
- Primary language
- TypeScript
- FA score
- Insufficient evidence
At a glance
- How it runs
- Works with
- Universal · cross-platformCodex · Claude Code
- Cost
- Free software; you pay for model usage
- Setup effort
- Medium · a few setup steps
- You'll need
- Typical use
- A developer who repeatedly follows a question, reproduction, analysis, planning, implementation, and testing loop for bugs can capture that sequence as a reusable workflow.
- Not a fit if
- Users unwilling to install and authenticate Claude Code, Codex, or Cursor CLIs
- Developers who need a graphical interface instead of a terminal
- Source review
- Insufficient evidence 8 safety controls not found
What does this agent do, and when should you use it?
CodeMachine is an open-source orchestration tool for AI coding engines operated through the command line. It captures development processes such as bug investigation, analysis, planning, implementation, and testing as workflows that can be run again. The project uses the headless scripting modes offered by AI coding engines, explicitly naming Claude Code, Codex, and Cursor. CodeMachine launches those engines through a CLI, supplies arguments and flags, passes context, and coordinates agents. It supports interactive-to-autonomous orchestration patterns, parallel work, and long-running workflows, but the supplied material does not document the workflow format, a workflow launch command, or where outputs are written.
A user defines a workflow once, then CodeMachine executes it through the headless scripting mode of AI coding engines. It spawns engines such as Claude Code, Codex, and Cursor via CLI, passes the applicable arguments and flags, and passes context between stages. A workflow can assign separate tasks to separate agents, allow those agents to communicate and collaborate, and run multiple agents in parallel. The default Ali Workflow Builder is presented as an interactive way to create a first workflow. The README describes persistence for workflows lasting hours or days, but does not disclose its storage implementation, workflow-definition syntax, or output-file conventions.
- A developer who repeatedly follows a question, reproduction, analysis, planning, implementation, and testing loop for bugs can capture that sequence as a reusable workflow.
- An engineering team that wants separate AI coding agents to handle research, design, coding, and review can coordinate those tasks in one workflow.
- A developer who wants agents to work on different portions of a workflow at the same time, instead of manually advancing each session in sequence, can use parallel execution.
- A user running coding automation for hours or days without continuously supervising the session can evaluate its long-running workflow support.
- A user who wants to build an initial automation flow interactively can start with the default Ali Workflow Builder.
How do you install or deploy this agent?
Install globally with: npm i -g codemachine. The supplied material does not specify a Node.js version, supported operating systems, required credentials, or how to install the target AI coding CLIs. After installation, an AI coding engine with a headless scripting mode is also required; Claude Code, Codex, and Cursor are explicitly named, but no copyable command for creating or launching a workflow is provided.
How do you use this agent?
Define a workflow in CodeMachine, or use the default Ali Workflow Builder to create a first workflow interactively. CodeMachine then invokes an AI coding engine with headless scripting support through its CLI and handles execution, context passing, and agent coordination. The supplied material contains no workflow configuration example, engine-authentication setup, workflow start command, or failure-recovery command, so those operational details are not established.
What are this agent's strengths and limitations?
- It is an orchestration layer for coding workflows that invokes AI coding engines through their headless CLI scripting modes.
- It explicitly supports task assignment across agents, agent communication, and parallel execution for decomposed development work.
- It centralizes context passing and dynamic-context management in the workflow infrastructure.
- It includes Ali Workflow Builder as a default interactive starting point for creating a workflow.
- Core execution depends on external AI coding CLIs that offer headless scripting; their installation, authentication, and version compatibility are not documented here.
- No workflow syntax, configuration-file structure, first-run command, or complete example is provided, adding adoption uncertainty.
- Long-running persistence is claimed, but the state location, recovery behavior, and operational limits are not described.
- The material does not specify the format or destination of code changes, test results, or other workflow outputs.
How does this agent compare with similar options?
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| CodeMachine This agent | Insufficient evidence | CLIFree + model costs | ★ 2.5k | 8mo ago | TypeScript | Codex · Claude Code |
| Swarms Multi-Agent Framework | Insufficient evidence | Library / SDKFree + model costs | ★ 7.4k | 1d ago | Python | Claude Code · OpenAI API · Claude API |
| Chat On Steroids | 84 · Good | Desktop appFree + model costs | ★ 5.1k | 1d ago | TypeScript | ChatGPT |
| Sortie | 71 · Some gaps | CLIFree + model costs | ★ 216 | 2d ago | Go | Codex · Claude Code |
How does FollowAgents rate this agent?
Why each dimension lost points
Evidence only includes README, LICENSE, package.json, and CI workflows. There is no documentation on least privilege, user confirmation, data flow transparency, sensitive data handling, dependency security, external effects, rollback, or source attribution. Hence all trust criteria score 0.
No code or tests to assess self-consistency, dependency availability, or failure messages. Hence all reliability criteria score 0.
README describes target audience and scenarios but lacks specifics on capability boundaries, trigger precision, or environment fit. Hence all adaptability criteria score 0.
README and LICENSE exist, but install notes, naming stability, examples, known limitations, versioning/changelog, and maintenance responsibility are missing. Hence all convention criteria score 0.
No evidence of output usability, marginal value, or cost-benefit. Hence all effectiveness criteria score 0.
No evidence supporting claim traceability, cross-source corroboration, or fact-inference separation. Hence all verifiability criteria score 0.
- Not found in source: least-privilege scopingGrant only what the task needs: a dedicated account or read-only token, scoped to specific directories and repos.
- Not found in source: confirmation before actingTurn on (or add) a confirmation step before it acts, and try it in a sandbox or test environment before real data.
- Not found in source: data-flow disclosureWatch which external services it contacts (proxy or firewall logs) and keep sensitive data out until you know where it goes.
- Not found in source: sensitive-data handlingUse dedicated, low-privilege, revocable API keys — never production credentials — and keep secrets out of logs.
- Not found in source: dependency securityPin versions and run a dependency audit (npm audit, pip-audit) before installing; prefer running it in a container.
- Not found in source: disclosed external effectsEstablish which external systems it writes to, sends to or changes, and verify with test accounts or repos before production.
- Not found in source: rollback or recovery pathBack up first, or work on a git branch or snapshot, so its changes can be undone.
- Not found in source: verifiable attributionInstall from the official repo or registry and check the publisher and URL to avoid look-alike packages.
- Repository contains only marketing material, no source code or tests, making substantive review impossible.
- Publisher identity is unverified, and there is no evidence of maintenance or update path.
- Dependencies are not pinned, posing supply chain risks.
Maintain this project? Add the FARS badge to your README
[](https://followagents.com/en/agents/codemachine-cli)<a href="https://followagents.com/en/agents/codemachine-cli"><img src="https://followagents.com/badge/codemachine-cli.svg" alt="FARS review"></a>Recent releases
Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.
-
v0.8.0 🚀 v0.8.0 Nova BETA Edition
Release notes excerpt
Changelog 🚀 Release v0.8.0 - Nova BETA Edition Nova removes the limits on workflow complexity. Controllers can now run entire workflows autonomously on your behalf. Workflows pause mid-execution and resume later - even after crashes. A new import system lets you share workflows across teams. And a rebuilt terminal interface gives you real-time visibility into exactly what your agents are doing. 🚀 Major Transformations: - Orchestration - Autonomous mode with controller agents, pause/resume, crash recovery, MCP-based agent coordination …View on GitHub ↗ -
v0.7.0 v0.7.0 - Bun Migration & OpenTUI
Release notes excerpt
🚀 Release v0.7.0 - Bun Migration & OpenTUI 🚀 Major Transformations: Complete Bun Runtime Migration: Revolutionary shift from Node.js/pnpm to Bun runtime with massive performance improvements - 98% faster builds (114ms vs several seconds) - 60% faster startup time (~200ms vs ~500ms) - 50% less memory usage (~40MB vs ~80MB) - 78% faster test execution (3s vs 14s) - Unified runtime for development, testing, and production Modern OpenTUI Interface: Complete terminal UI overhaul using OpenTUI/SolidJS - Beautiful, responsive TUI with custom CodeMachine theme …View on GitHub ↗ -
v0.5.0 Release v0.5.0 - OpenCode Integration & Type Safety
Release notes excerpt
🎉 New Features - OpenCode CLI Integration: Added OpenCode as a first-class engine provider alongside Claude, Codex, Cursor, and CCR - Full registry-based provider pattern with JSON event streaming - Non-interactive execution via environment guardrails - XDG-compliant paths under OPENCODE_HOME (~/.codemachine/opencode) - Support for multiple authentication providers - Run with: codemachine opencode run or --engine opencode in workflows 🔧 Improvements - Type Safety: Replaced 'any' types with 'unknown' and proper type guards throughout telemetry system …View on GitHub ↗
FAQ
Is it limited to one model or coding product?
Do I need anything installed besides CodeMachine?
codemachine installation. The exact setup requirements are not given.