fusion-3
Fusion full panel — Opus 4.8 + GPT-5.5 + Gemini 3.1 Pro in parallel, judged by Opus 4.8…
Fusion panel of Opus 4.8 + GPT-5.5 in parallel, judged by Opus 4.8 (opus4.8-gpt5.5)
$ npx -y skills add duolahypercho/fusion-fable --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
/fusion-gpt5.5Context preview
What this command does when you run it.
Fusion panel of Opus 4.8 + GPT-5.5 in parallel, judged by Opus 4.8 (opus4.8-gpt5.5)
description: Fusion panel of Opus 4.8 + GPT-5.5 in parallel, judged by Opus 4.8 (opus4.8-gpt5.5) argument-hint: <your question>
Invoke the **fusion** skill on the task below, forcing the `opus4.8-gpt5.5` panel: Opus 4.8 (Agent subagent) and GPT-5.5 (via `codex exec`) answer the SAME prompt IN PARALLEL, each independently with web + bash and neither seeing the other's work → Opus 4.8 judges both answers → Opus writes the final answer grounded in the analysis.
Follow the skill's SKILL.md exactly (fan out in parallel → judge → grounded final answer) and present the standard sections (Consensus / Contradictions / Partial coverage / Unique insights / Blind spots / Final answer). Use exactly one Opus 4.8 panelist and one GPT-5.5 panelist — do not add a Gemini panelist or a second Opus run. Pass the task verbatim to both; no "lenses". If the `codex` CLI is not installed, stop and say so rather than silently downgrading to the opus4.8-4.8 panel.
Task: $ARGUMENTS
Fuse a panel of frontier models into one Fable-tier answer. Fusion-Fable is a Claude Code skill that runs a hard question through a panel → judge pipeline.
Repo: duolahypercho/fusion-fable
Fusion full panel — Opus 4.8 + GPT-5.5 + Gemini 3.1 Pro in parallel, judged by Opus 4.8…
Fusion panel of Opus 4.8 + Gemini 3.1 Pro in parallel, judged by Opus 4.8…
Fusion panel of two independent Opus 4.8 runs, judged by Opus 4.8 (opus4.8-4.8)
OMC interview (interactive) → 3-round Opus 4.8 + GPT-5.5 fusion deepening → concise…