General-purpose agent driven by GPT-5.6 Luna at high reasoning effort.
General-purpose agent driven by GPT-5.6 Sol at high reasoning effort.
General-purpose agent driven by GPT-5.6 Sol at medium reasoning effort.
General-purpose agent driven by GPT-5.6 Terra at high reasoning effort.
Adversarial review of an implementation plan before work begins — invoke before exiting plan mode or committing to an approach. Pass the plan text or the path to a plan file.
Use when delegating work to subagents or Workflow agents and choosing which model and effort level to run them on — covers Claude vs GPT model strengths, cost dynamics, effort levels, and the mechanics of routing to GPT models through model-router.
Set up, verify, repair, or uninstall the model-router Claude/GPT routing gateway — binary install, CLIProxyAPI, Codex OAuth, OS service, and Claude Code settings wiring.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A shared tooling and knowledge layer for the AI alignment community. The repo behind alignment-hive.com.
As soft takeoff picks up, the alignment community needs shared infrastructure to keep pace. Alignment Hive aims to provide the benefits of scale that large labs have, through shared tooling and accumulated knowledge. AI tooling is moving fast, and it's hard to keep up with what's available and what works.
What's here:
Install Claude Code if you haven't already:
curl -fsSL https://claude.ai/install.sh | bash
curl -fsSL https://alignment-hive.com/install.sh | bash
This adds the plugin marketplace, installs the hive plugin, and walks you through data sharing preferences and project selection.
Session sharing requires an alignment-hive invite; everything else works without one. If you'd like an invite, email [email protected].
After installation open Claude Code in your project and run /hive:align.
To add the plugin marketplace without the install script (no CLI, no hive plugin), merge this into your ~/.claude/settings.json:
{
"extraKnownMarketplaces": {
"alignment-hive": {
"source": { "source": "github", "repo": "Crazytieguy/alignment-hive" },
"autoUpdate": true
}
}
}
Install a plugin by adding it to enabledPlugins in settings.json, e.g.:
{
"enabledPlugins": {
"mats@alignment-hive": true
}
}
| Plugin | Description | enabledPlugins key |
|---|---|---|
| hive | Tooling recommendations, session memory + sharing | Included in install script |
| mats | MATS fellow handbook, lit review, best practices | mats@alignment-hive |
| github-action | GitHub Action for autonomous @claude on issues and PRs | github-action@alignment-hive |
| llms-fetch-mcp | Documentation fetching with llms.txt support | llms-fetch-mcp@alignment-hive |
| remote-kernels | Cloud GPU machines with Jupyter kernels (RunPod, vast.ai, or Kubernetes) | remote-kernels@alignment-hive |
| model-router | GPT models as native Claude Code subagents via a local gateway (experimental) | model-router@alignment-hive |
Codex for Claude Code — Codex reviews and task delegation without leaving Claude Code — is maintained in a separate repo.
plugins/ — the Claude Code plugins abovepackages/web/ — alignment-hive.com, including the data-sharing backendpackages/hive-cli/ — CLI powering the hive plugincrates/remote-kernels/ — Rust binary behind the remote-kernels plugincrates/model-router/ — Rust gateway behind the model-router pluginFeedback and suggestions welcome. Open an issue or email [email protected].
npx claudepluginhub crazytieguy/alignment-hive --plugin model-routerMCP server for fetching documentation with llms.txt support
Cloud GPU instances with Jupyter kernels for AI/ML workloads
Share session learnings with the alignment research community
Autonomous operation with permission management for Claude Code
Set up Claude Code GitHub Action for autonomous issue and PR handling
Run AI models locally with Ollama - free alternative to OpenAI, Anthropic, and other paid LLM APIs. Zero-cost, privacy-first AI infrastructure.
TrueFoundry AI Gateway plugin — onboarding, model routing, MCP servers, prompts, Skills Registry workflows, observability, guardrails, and codebase migration. Works across Claude Code, Codex, and Cursor.
Run any model with an Anthropic- or OpenAI-compatible API (e.g. DeepSeek, GLM, Kimi, Qwen, MiniMax) — even your Codex subscription — as real Claude Code workflows, agent-team teammates, or one-shot subagents, driven exactly like native ones. Your main session's own auth is untouched (OAuth subscription or API key, either works); API-key providers bill the provider key via apiKeyHelper, while a Codex subscription bills through a local OAuth daemon — each worker receives its credential on demand, never through its env or argv. Requires the `cc-fleet` binary on PATH, installed separately.
Multi-LLM orchestration: delegate to Claude, Codex, and Gemini from a single interface with session management and async jobs.
Your ChatGPT/Codex subscription models in Claude Code's /model picker. A local gateway routes claude-codex-* models to OpenAI's Codex backend (via claude-code-proxy) and everything else to api.anthropic.com, so you switch between Claude and GPT-5.x mid-session like any other model. No OpenAI API key.
Teach Claude to detect, wire, and troubleshoot FreeRide — a local OpenAI-compatible gateway routing across free-tier providers (OpenRouter, Groq, NVIDIA NIM, Cloudflare Workers AI, HuggingFace).