メニュー
Model Selection Optimization (COST-003) — recommends optimal AI models for tasks given budget constraints, quality targets, and latency requirements. Computes the Pareto cost-quality frontier and supports mixed-model routing simulation.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Experiment orchestration framework with traffic allocation, statistical analysis, and early stopping detection. Use to test routing changes, model upgrades, and role assignments with Welch's t-test significance testing.
DEPRECATED — This skill is no longer maintained. Prometheus and Grafana infrastructure was never implemented. Use local JSON metrics analysis instead.
Scaffolds new SPEC-compliant agentic-engineers agents with a single call. Generates SKILL.md frontmatter, test scaffolds (TDD RED-phase), __init__.py, scripts/ layout, and DELEGATE/HANDBACK protocol templates. Use when creating any new automation agent, task handler, or operational tool. Validates role, model, effort, naming, and dependency graphs (circular dep detection) before writing files. Supports dry-run mode for planning without side effects.
Routine maintenance for Codex sessions: monitor queue state, close completed sub-agents, resume or escalate active work, and keep agent capacity available.
Automated cross-validation of protocol queue integrity. Scans all DELEGATEs/HANDBACKs, validates schema compliance, detects cycles, checks rate limits, generates compliance report. Enables self-referential protocol improvements.
Consolidates provider-specific AI costs into unified metrics across Anthropic, OpenAI, Google Gemini, GitHub Copilot, and Ollama. Enables apples-to-apples cost comparison and savings analysis.
| name | model-selection |
| description | Model Selection Optimization (COST-003) — recommends optimal AI models for tasks given budget constraints, quality targets, and latency requirements. Computes the Pareto cost-quality frontier and supports mixed-model routing simulation. |
| version | 1.0.0 |
| dependencies | [{"skill":"cost-budgeting","optional":false},{"skill":"cost-aggregation","optional":true}] |
| entry_points | ["scripts/model_selector.py"] |
| tests | ["tests/test_model_selector.py"] |
| coverage_minimum | 85 |
Recommends the optimal AI model for a task given cost, quality, and latency constraints.
ModelSelectorfrom scripts.model_selector import ModelSelector
selector = ModelSelector()
recommend_model(task_type, input_tokens, output_tokens, constraints=None)Recommend the single best model for a task.
Parameters:
task_type (str): Task category (code_review, documentation, security_audit, general, etc.)input_tokens (int): Expected input token countoutput_tokens (int): Expected output token countconstraints (dict, optional):
max_cost (float): Maximum cost per task in USDquality_target (float): Minimum quality score [0.0, 1.0]max_latency_sec (float): Maximum acceptable latency in secondsprovider_preference (list[str]): Ordered list of preferred providersReturns: dict with keys model, provider, estimated_cost, estimated_quality, estimated_latency_sec, reasoning, _selection_time_ms
recommend_batch(tasks)Recommend models for multiple tasks, tracking cumulative cost.
cost_quality_frontier(task_type, input_tokens, output_tokens, providers=None)Compute the Pareto-optimal cost/quality frontier.
simulate_model_mix(mix, daily_tasks, avg_tokens, task_type="general")Predict daily cost and quality for a hypothetical routing mix.
src/config/models.yamlcost_aggregator= parameter for richer cost estimates| Tier | Score Range | Example Models |
|---|---|---|
| mini | 0.52–0.67 | gpt-4o-mini |
| haiku | 0.55–0.70 | claude-haiku-4.5, gemini-flash |
| sonnet | 0.82–0.90 | claude-sonnet-4.5/4.6, gpt-4o |
| opus | 0.95–0.98 | claude-opus-4.8 |
This skill participates in the framework's continuous improvement cycle (see skill-improvement-feedback).
When you use model-selection during a task, include a skill_feedback entry in your HANDBACK to help improve it over time:
skill_feedback:
- skill_name: model-selection
effectiveness_score: 0.85 # required: 0.0–1.0
clarity_score: 0.90 # optional
coverage_gaps:
- "Specific scenario the skill did not address"
improvement_suggestions:
- "Concrete change that would have helped"
usage_context: "One sentence on how you used this skill"
Positive feedback is as valuable as critical feedback. Three or more feedback items for this skill automatically trigger an improvement task.