blank: open-source local-first agent harness fixes parallel sub-agent collisions and rate-limit halts
Big_Veterinarian7479 · reddit · 2026-11-08
A developer open-sourced blank (MIT), a local-first desktop & TUI agent harness addressing three pain points from daily-driving terminal coding agents:
- Parallel sub-agent file collisions: write-capable sub-agents (Scout/Architect/Critic, each on different models) run in isolated git worktrees so tasks never overwrite each other
- Rate-limit halts: register multiple API keys per provider for auto-rotation on 401/402/403/429, with fallback to secondary or local models
- CLI vs GUI trade-off: an Electron desktop app with split-view previews, Computer Use, ffmpeg/ComfyUI media pipelines and SSH over MCP, plus a full-screen terminal TUI
All state (sessions, checkpoints, memory.json, SKILL.md, masked audit logs) stays on disk in /.blank as plain files with zero telemetry, connecting directly to Ollama, LM Studio, vLLM, Claude, OpenAI, Gemini, DeepSeek and more.
More from coding & agent
- 17 open-source offensive AI security frameworks: Cairn hits 4.9k stars — evilsocket · 2026-11-08
- 'Find all the implicit state machines in my app' becomes a favorite AI prompt — rachittshah · 2026-11-08
- Claude Code sessions can message each other — demo shows agent-to-agent collaboration — dejanseo · 2026-11-08
- Qwen builds its own LLM: trains a 100k-parameter model in your browser, and narrates it — victormustar · 2026-11-08
- Dev open-sources Codemm: turn your own coding agent into a step-by-step tutor — gael55x · 2026-11-08
- Agent Lightning paper trains models through the same harness used at deployment — bibryam · 2026-11-08