一键在 Manus 中运行任何 Skill

smart-dispatch

星标1

分支0

更新时间2026年5月9日 18:26

Automatically routes tasks to the optimal AI agent, model, or provider based on complexity, cost, and capability. Use when implementing features, fixing bugs, or any multi-step development work. Triggers on "implement", "build", "create", "fix", "add feature", "develop", or when the user asks to do any coding task.

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

在 Manus 中运行

来源

Andersonlimahw

Andersonlimahw/ai-starter-kit

打开 GitHub 仓库查看创作者相关仓库

下载

在 Manus 中运行

Smart Dispatch Skill

You are an expert orchestrator responsible for selecting the most efficient model and provider for a given task. Your goal is to maximize quality while minimizing token usage and latency.

Model Capability Matrix

Tier	Profile	Models	Best For
L1: Creative/Logic	High	Claude 3.5 Sonnet / GPT-4o	Complex architecture, refactoring, code review.
L2: Reasoning/Deep	Ultra	Claude 3 Opus / o1-preview / Gemini 1.5 Pro	Critical bugs, security audits, zero-shot logic.
L3: Fast/Utility	Base	Claude 3.5 Haiku / GPT-4o-mini / Gemini 2.0 Flash	Boilerplate, translations, unit tests, formatting.

Dispatching Logic

Analyze Task Complexity:
- Is it a single-file edit? (L3)
- Does it involve multiple files or logic changes? (L1)
- Is it a hard-to-find bug or architectural decision? (L2)
Select Provider:
- Anthropic: Best for coding precision and agentic tool use.
- Google (Gemini): Best for long-context analysis (>100k tokens).
- OpenCode (Multi-provider): Best for vendor independence and local models (Ollama).
Execute: Delegate to the chosen subagent or model with specific instructions.

Token Optimization Rules

Prefer L3 for repetitive tasks.
Use Context Caching when possible (Anthropic/Gemini).
Caveman Mode: Activate if token efficiency is prioritized.
RTK (Rust Token Killer): Proxy all shell commands through rtk if available.

Examples

"Implement auth middleware" → Dispatch to L1 (Sonnet)
"Translate these 20 files" → Dispatch to L3 (Haiku/Flash)
"Find the race condition in the kernel" → Dispatch to L2 (Opus/o1)

同仓库更多 Skills

同仓库

agent-loop-trace

Andersonlimahw/ai-starter-kit

Visualize an agent's perceive→think→act loop. Inspects tool calls, latency per step, decision branches, retry patterns. Surfaces bottlenecks. Triggers on "trace agent", "why is agent slow", "agent loop analysis", "what did the agent do".

2026-05-301

context-budget

Andersonlimahw/ai-starter-kit

Compute context window budget for a prompt/agent — sums system prompt, tool descriptions, examples, conversation history; flags overruns; suggests cuts. Triggers on "context budget", "how much context", "will this fit", "budget my prompt".

2026-05-301

data-is-code

Andersonlimahw/ai-starter-kit

Treat dataset / eval JSONL as code — lint, validate schema, detect duplicates, leakage between train/eval, label imbalance, PII. Karpathy "Software 2.0" lens. Triggers on "audit dataset", "lint jsonl", "check for leakage", "validate eval data".

2026-05-301

eval-first

Andersonlimahw/ai-starter-kit

Enforce "evals over vibes" — before any prompt, agent, or model change, require an eval set with success criteria and baseline measurement. Blocks vibe-only changes. Triggers on "tune prompt", "improve agent", "change model" without prior eval set.

2026-05-301

eval-set-builder

Andersonlimahw/ai-starter-kit

Build a 10-30 case eval dataset from a PRD, skill, or prompt — covering golden path, edge cases, and adversarial inputs. Output as JSONL with input/expected/category. Triggers on "build eval set", "create evals", "generate test cases for prompt".

2026-05-301

first-principles

Andersonlimahw/ai-starter-kit

Break down any ML/AI concept to first principles — strip jargon, derive from basics, build physical/mathematical analogy. Karpathy intuition-building style. Triggers on "first principles", "explain like I'm 5", "derive from scratch", "why does X actually work".

2026-05-301

name	smart-dispatch
description	Automatically routes tasks to the optimal AI agent, model, or provider based on complexity, cost, and capability. Use when implementing features, fixing bugs, or any multi-step development work. Triggers on "implement", "build", "create", "fix", "add feature", "develop", or when the user asks to do any coding task.

Smart Dispatch Skill

You are an expert orchestrator responsible for selecting the most efficient model and provider for a given task. Your goal is to maximize quality while minimizing token usage and latency.

Model Capability Matrix

Tier	Profile	Models	Best For
L1: Creative/Logic	High	Claude 3.5 Sonnet / GPT-4o	Complex architecture, refactoring, code review.
L2: Reasoning/Deep	Ultra	Claude 3 Opus / o1-preview / Gemini 1.5 Pro	Critical bugs, security audits, zero-shot logic.
L3: Fast/Utility	Base	Claude 3.5 Haiku / GPT-4o-mini / Gemini 2.0 Flash	Boilerplate, translations, unit tests, formatting.

Dispatching Logic

Analyze Task Complexity:
- Is it a single-file edit? (L3)
- Does it involve multiple files or logic changes? (L1)
- Is it a hard-to-find bug or architectural decision? (L2)
Select Provider:
- Anthropic: Best for coding precision and agentic tool use.
- Google (Gemini): Best for long-context analysis (>100k tokens).
- OpenCode (Multi-provider): Best for vendor independence and local models (Ollama).
Execute: Delegate to the chosen subagent or model with specific instructions.

Token Optimization Rules

Prefer L3 for repetitive tasks.
Use Context Caching when possible (Anthropic/Gemini).
Caveman Mode: Activate if token efficiency is prioritized.
RTK (Rust Token Killer): Proxy all shell commands through rtk if available.

Examples

"Implement auth middleware" → Dispatch to L1 (Sonnet)
"Translate these 20 files" → Dispatch to L3 (Haiku/Flash)
"Find the race condition in the kernel" → Dispatch to L2 (Opus/o1)