wildwasser/opencode-agents
Code reviewこのページは AI により自動翻訳されています。技術的な詳細は原文をご確認ください。 原文を見る
Multi-agent AI development templates for opencode
概要
Multi-agent AI development templates for opencode. A ready-to-use template for setting up with opencode. Instead of a single AI assistant doing everything, work is delegated to specialized agents—each optimized for their role. - — Oscar stays lean, delegating heavy lifting to specialists - — Research, planning, and implementation are distinct phases - — Jester provides adversarial review for risky changes - — Independent tasks can run simultaneously For high-stakes decisions, run all three Jester variants in parallel and synthesize their feedback: - — Changing core abstractions, adding new patterns - — Changes touching >5 files or critical paths - — When you want multiple AI viewpoints on a problem - — When the team is stuck or going in circles 1. Oscar dispatches the same question to all three Jesters in parallel 2. Each Jester analyzes independently using their underlying model 3.
README
以下は原文 README の抜粋です。原文を見る。
opencode-agents
Multi-agent AI development templates for opencode.
What Is This?
A ready-to-use template for setting up multi-agent AI development workflows with opencode. Instead of a single AI assistant doing everything, work is delegated to specialized agents—each optimized for their role.
The Agents
Core Agents
| Agent | Role | Key Trait |
|---|---|---|
| Oscar | Orchestrator | Coordinates, delegates, synthesizes—never does the work himself |
| Scout | Researcher + Planner | Digs deep into codebases, creates actionable implementation plans |
| Ivan | Implementor | Writes code, runs tests, follows specs precisely |
| Jester | Truth-Teller (default) | Challenges assumptions, finds blind spots (called for risky changes) |
Jester Variants
| Agent | Model | Use Case |
|---|---|---|
| jester | Claude Opus | Default truth-teller |
| jester_opus | Claude Opus | Explicit Opus variant |
| jester_qwen | Qwen3 Coder | Code-focused analysis |
| jester_grok | Grok | Alternative perspective |
The Orchestrator Pattern
User Request
│
▼
Oscar ─────────────────────────────┐
│ │
├──→ Scout (research + plan) │
│ │ │
│ ├──→ Jester (challenge)│ ← optional
│ │ │
│ ▼ │
└──→ Ivan (implement) ──→ Done ◄─┘
Why this pattern?
- Context efficiency — Oscar stays lean, delegating heavy lifting to specialists
- Separation of concerns — Research, planning, and implementation are distinct phases
- Quality gates — Jester provides adversarial review for risky changes
- Parallel execution — Independent tasks can run simultaneously
Jester Consensus Pattern
For high-stakes decisions, run all three Jester variants in parallel and synthesize their feedback:
Oscar
│
├──→ @jester_opus ──┐
├──→ @jester_qwen ──┼──→ Synthesize → Decision
└──→ @jester_grok ──┘
When to use Jester Consensus:
- Major architectural decisions — Changing core abstractions, adding new patterns
- Risky refactors — Changes touching >5 files or critical paths
- Diverse perspectives needed — When you want multiple AI viewpoints on a problem
- Breaking ties — When the team is stuck or going in circles
How it works:
- Oscar dispatches the same question to all three Jesters in parallel
- Each Jester analyzes independently using their underlying model
- Oscar synthesizes the responses, looking for:
- Agreement — All three flag the same issue = high confidence
- Disagreement — Different concerns = explore each angle
- Unique insights — One Jester sees something others miss = investigate
Most of what any single Jester says is noise, but consensus across models is signal.
Installation
1. Install opencode
推奨ツール
別のキーワードを試すか、フィルタを外してください。
インストール
npx skillfish add wildwasser/opencode-agents