An open-source local dashboard called Claude Command Center (CCC) that manages multiple AI coding agent sessions across engines like Cursor, Claude Code, Codex, and Antigravity, providing cross-engine inbox, live status, git worktree spawning, and inter-agent group chat.
Been running 30+ AI coding agent sessions in parallel for about a year. Claude Code, Codex, Cursor (as of v4.7 last week), and Antigravity. The bottleneck stopped being the models pretty quickly — it became me, trying to keep track of which session was working, which was waiting, which had quietly finished an hour ago. So I built a local dashboard. CCC (Claude Command Center). Open source, MIT, vanilla JS + Python stdlib, no cloud, no account, no telemetry by default. **What it does:** * **Cross-engine inbox.** Cursor + Claude Code + Codex + Antigravity sessions in one view. Each engine is best at something; the dashboard lets me route by task instead of by app. * **Live session status at a glance.** Working, waiting, blocked on you, finished. Visible without opening each session. * **Worktree-aware spawning.** Every session can run in its own git worktree so parallel agents don't step on each other. * **Inter-agent group chat.** Sessions across engines can DM each other or chat in shared rooms. My most-used pattern: pin a Claude Sonnet session as the planner, spawn Cursor / Codex / Antigravity sub-sessions for execution, they coordinate via chat and only ping me when one of them needs me. * **Sees terminal sessions you didn't spawn.** Run `claude` or `codex` or `cursor-agent` from any terminal — they show up in the dashboard via on-disk transcript state. * **Bidirectional Cursor integration.** CCC-spawned Cursor sessions also appear in Cursor's native Agents Window. Both sides see the same work. * **Local models via Codex CLI.** Point `OPENAI_BASE_URL` at Ollama or LM Studio and CCC sees those sessions the same as cloud ones. Realistic floor for tool-use reliability is \~32B-class models. **Yes, this replaces the IDE for agent-driven work.** When agents write most of the code in parallel, you spend your time managing them, not typing. Your IDE becomes the tool you open when an agent flags something for manual touch — most of your day lives in CCC. **Install (macOS):** brew tap amirfish1/ccc brew install ccc Or directly from the git repository. There's also direct DMG to install - all in the comments. **What it isn't:** * Not a subscription bypass. Uses your existing auth. * Not cloud. Web dashboard at `localhost:8090`. Survives dashboard close. * Not magic. If you only run one session at a time, you don't need this. Happy to answer setup questions in the thread or DM.
Developers built an open-source web UI on top of the Cursor CLI that turns it into a multi-agent control panel, allowing users to run multiple Cursor agent sessions with separate workspaces, scheduling, and MCP config management from a browser-based cockpit.
Developer releases an open-source medieval-themed dashboard that visualizes multiple Claude Code agents as 2D village characters to simplify real-time tracking of parallel coding sessions.
The author built AgentsHive, a shared workspace that coordinates multiple AI agents (like Claude and Codex) into a collaborative product team with roles, memory, and routing, allowing solo builders to avoid manually juggling separate agent workflows.
Orca is an open-source AI orchestrator that lets developers run multiple CLI agents like Claude Code, Codex, and Gemini side-by-side across repositories on macOS, Windows, and Linux.
A developer discusses the challenge of coordinating multiple MCP-speaking AI agents (like Claude Code and Cursor) working on the same project, sharing their self-built open-source solution using a shared 'room' model inspired by IRC, and asking the community for patterns and opinions.