Agent Command Center: See which AI agents need you, then jump back in

Hi Product Hunt, I’m Takeshi, the maker of Agent Command Center. I built it because I use Claude Code and Codex in parallel. Starting another session is easy; remembering which session needs me, and where to return, is the part I kept losing. ACC reads the local session records those tools already write and turns them into one ledger. I open Resume, see what is running, waiting for me, asking, or stuck, and use Jump to return to the original Claude Code, Claude Desktop, Codex Desktop, or Codex CLI session. v0.2 is the largest update so far: Now lets me pin up to three current projects; notifications are opt-in; context handoff is confirmed before it writes; and Codex CLI, Claude Desktop Jump, and subagent details are included. A few boundaries matter to me: - The app has no LLM. - It makes no external network requests. - Classification is heuristic; when the evidence is not enough, it says “unknown.” - It is Apple Silicon only and paid-only, pay what you want from $9. If you finish one agent session before starting the next, you probably do not need ACC. It is for the moment when several sessions are moving and you catch yourself asking, “Where was that thread?” I use it every day. I would especially like to hear where your own multi-agent workflow becomes hard to track, and what boundary you would want documented more clearly.