Japanese engineer runs an AI agent org inside tmux: per-project teams, one human talking to the orchestrator
4310sy · x · 2026-10-16
Data/MLOps engineer uma-chan (uma-chan) shared his tmux-based multi-agent setup at #terminalnight:
- Org structure: one tmux session per project, panes running Claude Code, Codex, or other CLI agents, each staffed with progress-management, implementation, and review roles; he only talks to a "project oversight team" for status reports
- Key tmux mechanics: sessions/windows/panes for organization, dynamic pane management, send-keys to dispatch tasks, capture-pane to observe agents
- Cross-pane workflows: tasks dispatched via tmux send-keys -t %42; to track requests and replies he built the OSS tool tmux-a2a-postman, managing requests/responses as Markdown files with small commands, configs also in Markdown
Slides are publicly available.
More from coding & agent
- OpenAI stealth-testing mysterious internal model 'Aeon' in Codex desktop app and Dots — cedric_chee · 2026-10-16
- Vibe-coded iPhone Duo sniping bot built with Codex in 17 minutes, 41 tests passing — xiaohu · 2026-10-16
- Long-term multiplayer agent memory tool tops BEAM memory benchmark — CurieuxExplorer · 2026-10-16
- NativeCAD ships generative CAD over MCP with 177 tools, 94% of plain-English part requests pass — burhop · 2026-10-16
- Flux.ai imports STEP files and edits existing CAD geometry via natural language — burhop · 2026-10-16
- 50 tasks at once energizes me, says dev, as peer warns AI context-switching causes burnout — zack_overflow · 2026-10-16