AlphaClaw
The ultimate OpenClaw harness. Deploy in minutes. Stay running for months.
- Source repo
- chrysb/alphaclaw
- Stars
- ★ 1.5k
- Last updated
- 2d ago
- License
- MIT
- Primary language
- JavaScript
- FA score
- 53/100 · Major gaps
At a glance
- How it runs
- Works with
- Platform-specific
- Cost
- Free software; you pay for model usage
- Setup effort
- Medium · a few setup steps
- You'll need
- Typical use
- Developers who want to quickly deploy OpenClaw on Render or Railway without manual gateway setup.
- Not a fit if
- Teams that need an agent framework other than OpenClaw
- Developers doing local macOS development
- Teams unwilling to sync their workspace to GitHub
- Source review
- 53/100 · Major gaps
What does this agent do, and when should you use it?
AlphaClaw is a deployment and management framework for OpenClaw, designed to simplify the production setup and long-term operation of AI agents. It combines a user-friendly setup wizard, a self-healing watchdog, Git-backed rollback, and full browser-based observability, allowing operators to manage their agents without SSH or CLI. The framework provides a password-protected web dashboard that manages the OpenClaw gateway as a child process, with features for multi-agent management, channel orchestration (Telegram, Discord, Slack), Google Workspace integration, cron jobs, webhooks, file exploration, prompt hardening, and automatic Git sync. Targeting Docker/Linux deployments, it supports one-click deployment on Render and Railway, as well as self-hosting via npm or Docker.
AlphaClaw runs the OpenClaw gateway as a managed child process, monitoring its health via a watchdog that detects crashes, executes openclaw doctor --fix for auto-repair, and sends notifications. The Setup UI, built with Preact and Express, offers tabs for General, Browse, Usage, Cron, Nodes, Watchdog, Providers, Envars, and Webhooks, enabling configuration and management of agents, channels, cron jobs, environment variables, and webhook endpoints. It also injects anti-drift prompts (AGENTS.md, TOOLS.md) into the agent's system prompt, and automatically commits the workspace to GitHub on an hourly schedule. Users interact via the UI, CLI commands like alphaclaw start and alphaclaw git-sync, or the OpenAI-compatible /v1 proxy for programmatic access.
- Developers who want to quickly deploy OpenClaw on Render or Railway without manual gateway setup.
- Operations teams requiring self-healing capabilities for production OpenClaw deployments, with crash detection and automatic recovery.
- Teams managing multiple agent instances across Telegram, Discord, and Slack, needing a unified dashboard.
- Organizations that need to integrate Google Workspace services like Gmail, Calendar, and Drive into their agent workflows.
- Users who need scheduled tasks (cron jobs) and want to track usage and costs per agent.
- Those who require audit trails and version control for agent actions via automatic Git sync.
How do you install or deploy this agent?
Deploy via Render template, Railway template, or install locally with npm: npm install @chrysb/alphaclaw then npx alphaclaw start. For Docker, create a Dockerfile based on Node.js 22, install dependencies, set ALPHACLAW_ROOT_DIR, and expose port 3000. Set SETUP_PASSWORD and GITHUB_TOKEN environment variables as needed.
How do you use this agent?
After starting, visit the deployment URL and log in with SETUP_PASSWORD. Follow the guided onboarding to select a model, configure provider credentials, connect a GitHub repo, and pair channels. Use the UI tabs to manage gateway status, cron jobs, files, usage, and webhooks. CLI commands like alphaclaw start, alphaclaw git-sync -m "message", and alphaclaw telegram topic add are also available.
What are this agent's strengths and limitations?
- Self-healing watchdog detects crashes, runs
openclaw doctor --fix, and restarts the gateway automatically. - Browser-based management eliminates the need for SSH or manual config file editing, with file explorer and inline edits.
- Automatic Git sync with hourly commits ensures every agent action is version-controlled and auditable.
- Multi-channel orchestration with guided wizard for Telegram topics, Discord, and Slack pairings.
- One-click deployment templates for Render and Railway with minimal setup.
- Exclusively tied to OpenClaw; not reusable with other agent frameworks.
- Targets Docker/Linux; macOS local development is not supported.
- Security trade-offs include single setup password, auto-approval of pairings, and query-string tokens, which reduce OpenClaw's default hardening.
- Requires a GitHub token and at least 8GB RAM for Railway deployments (Trial plan may cause OOM crashes).
- Deployment depends on external platforms unless self-hosted with Docker.
How does this agent compare with similar options?
Compared to using OpenClaw directly, AlphaClaw provides a management UI and self-healing capabilities but sacrifices some of OpenClaw's security hardening for ease of use.
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| AlphaClaw This agent | 53 · Major gaps | Self-hosted serviceFree + model costs | ★ 1.5k | 2d ago | JavaScript | — |
| ClawMetry | 81 · Good | CLIFreemium | ★ 432 | 1d ago | Python | Codex · Claude Code |
| lazyagent Coding-Agent Monitor | 75 · Good | CLIFree | ★ 193 | 2d ago | Go | Codex · Claude Code |
| NVIDIA NemoClaw | 65 · Some gaps | CLIFree + model costs | ★ 23k | today | TypeScript | — |
How does FollowAgents rate this agent?
Why each dimension lost points
Evidence shows: README explicitly states AlphaClaw is a convenience wrapper for OpenClaw and lists security trade-offs (e.g., single SETUP_PASSWORD, auto-approval of pairings, query-string tokens). Built-in brute-force protection (exponential backoff lockout). Sensitive data handling: tokens stored in .env file with noted risks. Dependency security: few dependencies (express, http-proxy, openclaw, ws) but no vulnerability scanning or locked versions (uses ^ ranges). External effects: auto-repair, Git sync, cron installation, but no user confirmation mechanism. Rollback: Git backup and version management, but no explicit rollback procedure. Source attribution: unverified publisher, but README and package.json provide repository links. Deductions: lack of user confirmation, limited data flow transparency, insufficient dependency security.
Evidence shows: README and test files describe a consistent architecture and functionality. Tests cover bin commands, doctor functionality, config restoration, etc. Dependency availability: dependencies from npm, but no lock file or integrity checks. Failure messages: tests check error messages like port conflict and missing SETUP_PASSWORD. Deductions: dependency availability not fully verified, failure message coverage limited.
Evidence shows: README describes multiple deployment scenarios (Render, Railway, Docker, local) and provides environment variables and CLI commands. Capability boundaries: clearly states it is a wrapper for OpenClaw, not a replacement. Trigger precision: CLI commands and UI actions are clearly described. Environment fit: supports Docker/Linux, but explicitly states macOS local development is not supported. Deductions: trigger precision may not be precise for UI actions, environment fit limited.
Evidence shows: README is well-structured with features, quick start, architecture, environment variables, security notes, etc. Install notes: provides npm and Docker installation methods. Naming stability: package name and CLI commands consistent. Examples and FAQ: missing FAQ and detailed examples. Known limitations: clearly states platform limitations and security trade-offs. License: MIT license. Versioning and changelog: package.json has version number, but no changelog provided. Maintenance responsibility: no clear maintainer or contribution guidelines. Deductions: missing FAQ, changelog, and maintenance responsibility.
Evidence shows: Output usability: provides CLI and UI interfaces with clear output formats. Marginal value: as a wrapper for OpenClaw, provides additional management features like watchdog, Git sync, Web UI. Cost-benefit: free and open source, but requires deployment and maintenance costs. Deductions: marginal value may be limited as core functionality depends on OpenClaw.
Evidence shows: Some claims in README are supported by tests (e.g., test files), but no independent verification. Cross-source corroboration: no external verification. Fact-inference separation: README distinguishes feature descriptions and security trade-offs, but does not explicitly separate facts and inferences. Deductions: limited claim traceability, insufficient cross-source corroboration.
- Publisher identity is unverified; proceed with caution.
- Dependencies are not pinned, posing supply chain risks.
- Security trade-offs are explicit, but users must assess risks themselves.
- Lack of user confirmation mechanisms may introduce risks from automated actions.
- No changelog provided, making version updates opaque.
Recent releases
Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.
-
v0.9.37 AlphaClaw 0.9.37
Release notes excerpt
AlphaClaw 0.9.37 What's New * Updated model picker: Recommend Claude 5.5, GPT-6.1 Sol, GPT-6 Astra/Luna, and Gemini 3.8 Flash. Older models remain available. * Readable watchdog logs: Theme-friendly red errors and yellow warnings. * OpenClaw 2026.10.2: Promote the version tested in Railway beta. Fixes * Reliable gateway restarts: Wait for shutdown before replacement and preserve process ownership across restarts. * GPT-6.1 Sol authentication: Support Codex OAuth throughout onboarding and server-side validation. Contributors * @chrysbView on GitHub ↗ -
v0.9.36 AlphaClaw 0.9.36
Release notes excerpt
AlphaClaw 0.9.36 Fixes * Cron run history: Restore Recent runs, trends, and duration metrics after OpenClaw’s SQLite migration while retaining compatibility with older JSONL-based installations.View on GitHub ↗ -
v0.9.35 AlphaClaw 0.9.35
Release notes excerpt
AlphaClaw 0.9.35 What's New * OpenClaw 2026.9.3 support: Upgrade OpenClaw with GPT-6 Astra routing, pricing, thinking levels, updated SecretRef and supervisor contracts, and bootstrap compatibility. (#126) * Safe config upgrades: Automatically migrate older OpenClaw configurations before startup, preserve explicit settings, validate the result, and restore the original configuration if migration fails. (#126) * Modern runtime: Move to Express 5.2.1 and require Node.js 24.16+ or 26.1+ for supported installations. (#126) Fixes …View on GitHub ↗
FAQ
Is AlphaClaw free?
How secure is it?
SETUP_PASSWORD, and AlphaClaw includes brute-force protection. However, be aware of query-string tokens in webhook URLs and the gateway token stored in .env.What happens if the gateway crashes?
openclaw doctor --fix, and restarts the gateway, notifying via Telegram/Discord/Slack.