fable-advisor: an open-source orchestrator routing Opus, GPT-6 and Grok in Claude Code
daniel_mac8 · x · 2026-09-30
Daniel Mac argues GPT-6.1 Sol doesn't top Opus or Sonnet 5.5 on intelligence per Artificial Analysis but is Pareto-optimal on efficiency, making it ideal as a subagent. His free open-source fable-advisor plugin (1.1k stars) exploits this:
- Claude Code lets each subagent run a different model, and the session itself run on another — the plugin implements the architect pattern: the session model owns requirements, decomposition, specs, and verification
- Lane routing: routine tasks go to GPT-6 Luna (effort max via Codex CLI), high-complexity tasks to GPT-6 Sol (effort high)
- Finished work gets a clean-context review from Fable 5.1 before anything ships
More from coding & agent
- Dev builds Rez-inspired music rail shooter with Sonnet then Opus, playable in browser — AIandDesign · 2026-09-30
- Ramp demos agentic buying: Codex agent picks wine via Stripe MCP and pays with a Ramp card over MPP — jeff_weinstein · 2026-09-30
- AI agent starts reverse engineering NVIDIA drivers to run game benchmarks — mgostIH · 2026-09-30
- ModRetro console ships with blank cartridges you fill by coding games with Codex — OpenAIDevs · 2026-09-30
- Beam Studio launches on Bittensor Subnet 105 as a coordination layer for AI agents — markjeffrey · 2026-09-30
- Open-source project brings LoRA training to 50+ models — no3us · 2026-09-30