ClawTeam-OpenClaw and openclaw-operator
About ClawTeam-OpenClaw
win4r/ClawTeam-OpenClaw
ClawTeam fork fully adapted for OpenClaw — multi-agent swarm coordination with OpenClaw as the default agent
Leverages **git worktrees and tmux windows** to provide true process isolation per agent (no container overhead), with agents coordinating through filesystem and optional ZeroMQ P2P messaging. Integrates with any CLI-based coding agent—OpenClaw, Claude Code, Codex, Cursor—via auto-injected `clawteam` commands that agents call directly from their prompts to spawn workers, manage task queues, and report results.
About openclaw-operator
DannyMac180/openclaw-operator
OpenClaw Operator gives coding agents like Codex and Claude Code the context and playbooks needed to set up, validate, and troubleshoot a local OpenClaw installation.
Provides structured playbooks and validation checklists that help coding agents inspect OpenClaw state before editing, make minimal reversible changes, and verify results—deploying as reusable skills compatible with Codex, Claude Code, and OpenClaw's native agent framework. Architecture emphasizes safe operational workflows: inspect-before-edit, narrow diffs, security hygiene checks, and health validation rather than automation magic. Integrates directly into agent-specific skill directories (`.agents/skills/`, `.claude/skills/`, `~/.openclaw/skills/`) and accepts plain-English prompts to handle common tasks like skill creation, provider wiring, and cron job setup.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work