Agent skill
task-classifier
Classify task complexity (1-10) to route to optimal model. Use when starting any non-trivial task to determine resources and approach. Triggers include: 'classify', 'complexity', 'what model', 'how complex'.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/task-classifier
SKILL.md
Skill: Task Classifier
ultrathink - Take a deep breath. We're not here to write code. We're here to make a dent in the universe.
The Vision
Classification should make the path forward feel inevitable.
Your Work, Step by Step
- Assess scope: Files, modules, and integrations.
- Assess risk: Security, breakage, rollback.
- Assess ambiguity: Unknowns and trade-offs.
- Assign complexity: Recommend models and iterations.
Ultrathink Principles in Practice
- Think Different: Avoid defaulting to mid complexity.
- Obsess Over Details: Map true blast radius.
- Plan Like Da Vinci: Evaluate before judging.
- Craft, Don't Code: Keep reasoning concise.
- Iterate Relentlessly: Reclassify as scope changes.
- Simplify Ruthlessly: Choose the simplest valid path.
Purpose
Classify task complexity (1-10) to route to optimal model.
Complexity Matrix
| Score | Complexity | Examples | Recommended Model |
|---|---|---|---|
| 1-2 | Trivial | Typo fix, comment update | MiniMax-lightning |
| 3-4 | Simple | Add field, simple function | MiniMax M2.1 |
| 5-6 | Medium | New API endpoint, refactor | Sonnet → Codex/Gemini |
| 7-8 | High | Multi-file feature, migration | Opus → Sonnet → CLIs |
| 9-10 | Exceptional | Architecture redesign, security fix | Opus (thinking) |
Classification Criteria
Technical Factors
- Files affected (1 = trivial, 10+ = exceptional)
- Cross-module dependencies
- Database schema changes
- External API integrations
Risk Factors
- Security implications (auth, payments, data)
- Breaking change potential
- Rollback difficulty
Cognitive Factors
- Novel problem (no existing patterns)
- Ambiguity in requirements
- Trade-off decisions required
Output Format
{
"complexity": 7,
"reasoning": "Multi-file feature with database changes and auth integration",
"recommended_model": "opus",
"delegation": ["codex:security", "gemini:integration-tests"],
"estimated_iterations": 12,
"risk_level": "medium"
}
Model Routing
| Complexity | Primary | Secondary | Fallback |
|---|---|---|---|
| 1-2 | MiniMax-lightning | - | - |
| 3-4 | MiniMax M2.1 | - | - |
| 5-6 | Sonnet | Codex/Gemini | MiniMax |
| 7-8 | Opus | Sonnet + CLIs | MiniMax |
| 9-10 | Opus (thinking) | Codex | Gemini |
Iteration Limits
| Model | Max Iterations |
|---|---|
| Claude (Sonnet/Opus) | 15 |
| MiniMax M2.1 | 30 |
| MiniMax-lightning | 60 |
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?