From 2ce69710e335afcaa82769ee55cdcb1e5de3efa0 Mon Sep 17 00:00:00 2001 From: YeonGyu-Kim Date: Tue, 17 Mar 2026 15:43:08 +0900 Subject: [PATCH] docs: sync agent-model-matching guide with actual fallback chains - Metis: add missing GPT-5.4 high as 2nd fallback - Hephaestus: add GPT-5.4 (Copilot) fallback, was incorrectly listed as Codex-only - Oracle: add opencode-go/glm-5 as last fallback - Momus: add opencode-go/glm-5 fallback, note xhigh variant - Atlas: add GPT-5.4 medium as 3rd fallback - Sisyphus: add Kimi K2.5 (moonshot providers) in chain - Sisyphus-Junior: add missing agent to Utility Runners section - GPT Family table: merge duplicate GPT-5.4 rows - Categories: add missing opencode-go intermediate fallbacks for visual-engineering, ultrabrain, quick, unspecified-low/high, writing --- docs/guide/agent-model-matching.md | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/docs/guide/agent-model-matching.md b/docs/guide/agent-model-matching.md index 4f13c2664..ba2130bf0 100644 --- a/docs/guide/agent-model-matching.md +++ b/docs/guide/agent-model-matching.md @@ -64,8 +64,8 @@ These agents have Claude-optimized prompts — long, detailed, mechanics-driven. | Agent | Role | Fallback Chain | Notes | | ------------ | ----------------- | -------------------------------------- | ------------------------------------------------------------------------------------------------- | -| **Sisyphus** | Main orchestrator | Claude Opus → opencode-go/kimi-k2.5 → K2P5 → GPT-5.4 → GLM-5 → Big Pickle | Claude-family first. GPT-5.4 has dedicated prompt support. Kimi/GLM as intermediate fallbacks. | -| **Metis** | Plan gap analyzer | Claude Opus → opencode-go/glm-5 → K2P5 | Claude preferred. Uses opencode-go for reliable GLM-5 access. | +| **Sisyphus** | Main orchestrator | Claude Opus → opencode-go/kimi-k2.5 → K2P5 → Kimi K2.5 → GPT-5.4 → GLM-5 → Big Pickle | Claude-family first. GPT-5.4 has dedicated prompt support. Kimi available through multiple providers. | +| **Metis** | Plan gap analyzer | Claude Opus → GPT-5.4 → opencode-go/glm-5 → K2P5 | Claude preferred. GPT-5.4 as secondary before GLM-5 fallback. | ### Dual-Prompt Agents → Claude preferred, GPT supported @@ -74,7 +74,7 @@ These agents ship separate prompts for Claude and GPT families. They auto-detect | Agent | Role | Fallback Chain | Notes | | -------------- | ----------------- | -------------------------------------- | -------------------------------------------------------------------- | | **Prometheus** | Strategic planner | Claude Opus → GPT-5.4 → opencode-go/glm-5 → Gemini 3.1 Pro | Interview-mode planning. GPT prompt is compact and principle-driven. | -| **Atlas** | Todo orchestrator | Claude Sonnet → opencode-go/kimi-k2.5 | Claude first, opencode-go as the current fallback path. | +| **Atlas** | Todo orchestrator | Claude Sonnet → opencode-go/kimi-k2.5 → GPT-5.4 | Claude first, opencode-go as intermediate, GPT-5.4 as last resort. | ### Deep Specialists → GPT @@ -82,9 +82,9 @@ These agents are built for GPT's principle-driven style. Their prompts assume au | Agent | Role | Fallback Chain | Notes | | -------------- | ----------------------- | -------------------------------------- | ------------------------------------------------ | -| **Hephaestus** | Autonomous deep worker | GPT-5.3 Codex only | No fallback. Requires GPT access. The craftsman. | -| **Oracle** | Architecture consultant | GPT-5.4 → Gemini 3.1 Pro → Claude Opus | Read-only high-IQ consultation. | -| **Momus** | Ruthless reviewer | GPT-5.4 → Claude Opus → Gemini 3.1 Pro | Verification and plan review. | +| **Hephaestus** | Autonomous deep worker | GPT-5.3 Codex → GPT-5.4 (Copilot) | Requires GPT access. GPT-5.4 via Copilot as fallback. The craftsman. | +| **Oracle** | Architecture consultant | GPT-5.4 → Gemini 3.1 Pro → Claude Opus → opencode-go/glm-5 | Read-only high-IQ consultation. | +| **Momus** | Ruthless reviewer | GPT-5.4 → Claude Opus → Gemini 3.1 Pro → opencode-go/glm-5 | Verification and plan review. GPT-5.4 uses xhigh variant. | ### Utility Runners → Speed over Intelligence @@ -95,6 +95,7 @@ These agents do grep, search, and retrieval. They intentionally use the fastest, | **Explore** | Fast codebase grep | Grok Code Fast → opencode-go/minimax-m2.5 → MiniMax Free → Haiku → GPT-5-Nano | Speed is everything. Fire 10 in parallel. | | **Librarian** | Docs/code search | opencode-go/minimax-m2.5 → MiniMax Free → Haiku → GPT-5-Nano | Doc retrieval doesn't need deep reasoning. | | **Multimodal Looker** | Vision/screenshots | GPT-5.4 → opencode-go/kimi-k2.5 → GLM-4.6v → GPT-5-Nano | Uses the first available multimodal-capable fallback. | +| **Sisyphus-Junior** | Category executor | Claude Sonnet → opencode-go/kimi-k2.5 → GPT-5.4 → Big Pickle | Handles delegated category tasks. Sonnet-tier default. | --- @@ -119,8 +120,7 @@ Principle-driven, explicit reasoning, deep technical capability. Best for agents | Model | Strengths | | ----------------- | ----------------------------------------------------------------------------------------------- | | **GPT-5.3 Codex** | Deep coding powerhouse. Autonomous exploration. Required for Hephaestus. | -| **GPT-5.4** | High intelligence, strategic reasoning. Default for Oracle. | -| **GPT-5.4** | Strong principle-driven reasoning. Default for Momus and a key fallback for Prometheus / Atlas. | +| **GPT-5.4** | High intelligence, strategic reasoning. Default for Oracle, Momus, and a key fallback for Prometheus / Atlas. Uses xhigh variant for Momus. | | **GPT-5-Nano** | Ultra-cheap, fast. Good for simple utility tasks. | ### Other Models @@ -166,14 +166,14 @@ When agents delegate work, they don't pick a model name — they pick a **catego | Category | When Used | Fallback Chain | | -------------------- | -------------------------- | -------------------------------------------- | -| `visual-engineering` | Frontend, UI, CSS, design | Gemini 3.1 Pro → GLM 5 → Claude Opus | -| `ultrabrain` | Maximum reasoning needed | GPT-5.4 → Gemini 3.1 Pro → Claude Opus | +| `visual-engineering` | Frontend, UI, CSS, design | Gemini 3.1 Pro → GLM 5 → Claude Opus → opencode-go/glm-5 → K2P5 | +| `ultrabrain` | Maximum reasoning needed | GPT-5.4 → Gemini 3.1 Pro → Claude Opus → opencode-go/glm-5 | | `deep` | Deep coding, complex logic | GPT-5.3 Codex → Claude Opus → Gemini 3.1 Pro | | `artistry` | Creative, novel approaches | Gemini 3.1 Pro → Claude Opus → GPT-5.4 | -| `quick` | Simple, fast tasks | Claude Haiku → Gemini Flash → GPT-5-Nano | -| `unspecified-high` | General complex work | Claude Opus → GPT-5.4 (high) → GLM 5 → K2P5 | -| `unspecified-low` | General standard work | Claude Sonnet → GPT-5.3 Codex → Gemini Flash | -| `writing` | Text, docs, prose | Gemini Flash → Claude Sonnet | +| `quick` | Simple, fast tasks | Claude Haiku → Gemini Flash → opencode-go/minimax-m2.5 → GPT-5-Nano | +| `unspecified-high` | General complex work | Claude Opus → GPT-5.4 → GLM 5 → K2P5 → opencode-go/glm-5 → Kimi K2.5 | +| `unspecified-low` | General standard work | Claude Sonnet → GPT-5.3 Codex → opencode-go/kimi-k2.5 → Gemini Flash | +| `writing` | Text, docs, prose | Gemini Flash → opencode-go/kimi-k2.5 → Claude Sonnet | See the [Orchestration System Guide](./orchestration.md) for how agents dispatch tasks to categories.