hive
claudeTooling recommendations and session sharing for alignment researchers
3 skills · 1 command · 1 agent · hooks
Non-Claude models (GPT, Grok, Muse) as native Claude Code subagents via a local routing gateway
Plugin model-router from marketplace alignment-hive in Crazytieguy/alignment-hive. Commands read from the repo's manifest — yesterday.
delegating-to-modelsThis skill should be read before delegating work to subagents or Workflow agents. Covers picking a model from the Claude, GPT, and optionally Grok and Muse families (strengths, cost dynamics), setting the effort level, prompt hygiene per family, combining independent runs, and the mechanics of routi
setupSet up, configure, verify, repair, or uninstall model-router, the local gateway that routes Claude Code to non-Claude models, including any of its config options — binary install, CLIProxyAPI, Codex OAuth or OpenAI API key, optional Grok (xAI), Muse (Meta) and open-weights models, Ultrafast GPT mode
adversarial-code-reviewerAdversarial code review by GPT-6.1 Sol. State the scope (uncommitted changes, a commit, or a base ref) and the repo or worktree path, plus brief background the diff can't show. The reviewer derives its own attack angles and has its materiality bar, output format, and read-only rules built in — don't
gpt-6-astra-fastGeneral-purpose agent driven by GPT-6 Astra on Codex's fast tier, 1.5–2x faster for 2.5x the usage. Read the model-router:delegating-to-models skill before delegating to it.
gpt-6-astraGeneral-purpose agent driven by GPT-6 Astra. Read the model-router:delegating-to-models skill before delegating to it.
gpt-6-luna-fastGeneral-purpose agent driven by GPT-6 Luna on Codex's fast tier, 1.5–2x faster for 2.5x the usage. Read the model-router:delegating-to-models skill before delegating to it.
gpt-6-lunaGeneral-purpose agent driven by GPT-6 Luna. Read the model-router:delegating-to-models skill before delegating to it.
gpt-6.1-sol-fastGeneral-purpose agent driven by GPT-6.1 Sol on Codex's fast tier, 1.5–2x faster for 2.5x the usage. Read the model-router:delegating-to-models skill before delegating to it.
gpt-6.1-solGeneral-purpose agent driven by GPT-6.1 Sol. Read the model-router:delegating-to-models skill before delegating to it.
gpt-6.1-terra-fastGeneral-purpose agent driven by GPT-6.1 Terra on Codex's fast tier, 1.5–2x faster for 2.5x the usage. Read the model-router:delegating-to-models skill before delegating to it.
gpt-6.1-terraGeneral-purpose agent driven by GPT-6.1 Terra. Read the model-router:delegating-to-models skill before delegating to it.
plan-reviewerAdversarial review of an implementation plan by GPT-6.1 Sol — invoke before committing to an approach. Pass the path to the plan file (or the plan text), plus brief background the plan doesn't state: the goal and any decisions the user has already fixed. The reviewer derives its own critique angles
Tooling recommendations and session sharing for alignment researchers
3 skills · 1 command · 1 agent · hooks
Deprecated: superseded by Claude Code's built-in auto mode. Autonomous operation with permission management.
2 skills · hooks
Debriefs of completed agent work
1 skill · hooks
Literature review pipeline (arXiv, Semantic Scholar, LessWrong)
1 skill · 1 agent
Cloud GPU instances with Jupyter kernels for AI/ML workloads
1 skill · 1 MCP server
Deprecated: no longer maintained. Set up Claude Code GitHub Action for autonomous issue and PR handling.
1 skill
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.