itr8 puts Claude, GPT, Gemini and Grok in one group chat with a rich agent canvas
cantrell · x · 2026-11-09
Developer cantrell ran few-shot testing with itr8, an "AI group chat superapp":
- Brings models from Anthropic, OpenAI, Google, xAI and OpenRouter into a single group chat where they divide work, debate, and review each other's output alongside humans
- Each agent gets a persona and skills (security reviewer, patent attorney, skeptical VC), with new specialists joining as work evolves
- Ships a rich interactive canvas: annotated docs, code editors, kanban boards, image/video editing, spreadsheets and diagrams—all editable by both agents and humans
- The pitch leverages models' "jagged" capabilities: combining them beats any single model
More from coding & agent
- Dev slams Codex app: cloud-to-local handoff is 'completely broken' — adhamelarabawy · 2026-11-09
- Scott Hanselman open-sources sandbox-mcp: disposable Windows VMs for AI agents — davemccollough · 2026-11-09
- PM runs 5-6 Claude Code agents in parallel for 24+ hours with 5 no-code principles — PawelHuryn · 2026-11-09
- Dev lets three Claude agents live in a virtual town; adding personalities makes them far more interesting — johnnypen2445 · 2026-11-09
- A single AI prompt turns agents into a product design review team that screenshots and fixes UI issues — vista8 · 2026-11-09
- Stanford and DeepMind evolve stale benchmarks into hard tests with TURTLE, targeting 50% solver success — rohanpaul_ai · 2026-11-09