OpenClaw Plugin for Claude Code and Codex Orchestration
Plan approval and session persistence for AI agents via Telegram or Discord chat.
Feels like Codespaces for LLM-driven coding: each session gets its own git worktree plus a browser xterm and a virtual mobile keyboard. The background 'shepherd' that preserves PTY sessions across server restarts and the single Go binary embedding the React frontend are neat, practical engineering decisions — just remember you need the Claude/Codex CLIs and to lock down access when exposing the server.
Developers and teams who want self-hosted, persistent LLM-powered coding sessions tied to GitHub repositories
Some neat features: - Web UI with mobile friendly buttons (arrows, enter, etc.) - Separate background task for agents (so the server can restart without destroying in progress sessions) - Github integration - Notifications when user intervention is required
Hope y'all like it!
Plan approval and session persistence for AI agents via Telegram or Discord chat.
One Dockerfile to rule Claude Code, OpenCode, and Codex with persistent sessions.
Session hooks auto-inject product context—no more re-briefing Claude every new session.
Persistent agent sessions accessible from anywhere — SSH for AI workflows done right.
Claude agent that learns from mistakes via structured signal tracing—interesting approach, unclear real-world ROI.
Claude Code memory server that auto-reads OAT tokens, routes LLM ops to local or cloud models.