本文へ移動
cccskills
無料GitHub で公開

model-selection

Choose and tune which AI model a custom agent, subagent, or squad node should use. USE FOR: picking a model by task capability tier (general-purpose coding/writing, fast/cheap, deep reasoning/debugging, multimodal); efficiency-first vs. capability-first starting strategies; tuning an effort/reasoning-effort parameter instead of switching models; call cadence vs. provider prompt-cache TTL; cost-tier ranking and cost ceilings; cross-provider or cross-model-family diversity for independent/adversarial review; per-harness model field syntax and resolution order (GitHub Copilot CLI, Claude Code, VS Code custom agents); extending this framework to a new harness. DO NOT USE FOR: authoring the rest of an agent's frontmatter or prompt body (use copilot-cli-custom-agents, claude-code-custom-agents, or the target harness's own skill); designing squad topology (use agent-graph-orchestration); SKILL.md authoring (use create-skill).

日本語の概要は準備中です。原文の説明を表示しています。

NetFabric/NetFabric.Numericsスター 36(GitHub 36 + cccskills 0)
連携ツールGitHub

実行スコア 未検証

インストール方法を見る

含まれるファイル（4）

  • SKILL.md4.4 KB
  • references/cache-and-cost.md4.0 KB
  • references/harness-mapping.md3.2 KB
  • references/selection-criteria.md4.1 KB

SKILL.md（原文）

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Model Selection

Every custom agent, subagent, or squad node needs an explicit model choice — never leave it unset on the assumption that an inherited default is adequate for a specialist role. The right model balances five criteria, then gets mapped onto whatever field syntax the target harness actually accepts.

Anatomy

CriterionQuestion it answers
Capability tierHow complex is the reasoning this agent's task requires?
SpeedHow latency-sensitive is whoever is waiting on this agent?
CostWhat's the budget, and is this agent bounded by a caller's own cost ceiling?
Context & modalityDoes the task need a large context window, or image/visual input?
EffortCan the same model be tuned cheaper or faster instead of switching models entirely?

Workflow

  1. Establish task criteria for this specific agent — capability need, latency sensitivity, cost budget, context/multimodal needs → references/selection-criteria.md
  2. Pick a starting strategy: efficiency-first (start cheap/fast, upgrade only if evals show a gap) or capability-first (start at the top, optimize down later) → references/selection-criteria.md
  3. Prefer tuning an effort/reasoning-effort parameter over switching models when the harness and model support it — it trades intelligence for latency/cost within the same model family, a cheaper lever than a full model swap → references/selection-criteria.md
  4. If this agent is one node of a multi-agent squad, or needs independent/adversarial review, apply call-cadence vs. provider prompt-cache TTL, cost-tier ranking/ceilings, and cross-provider diversity → references/cache-and-cost.md
  5. Ask which harness targets this agent (if not already established), then apply that harness's exact model field name, value shape, and resolution order → references/harness-mapping.md
  6. Verify the chosen model name is still current against the provider's live docs before finalizing — model lineups rotate every few months; never trust a memorized or previously-seen model name as still accurate → references/selection-criteria.md
  7. Run markdown-best-practices over any new or edited Markdown produced while documenting the choice (design notes, agent files' surrounding docs)

Reference Files

FileLoad When
references/selection-criteria.mdPicking a capability tier for a task, choosing an efficiency-first vs. capability-first starting point, tuning an effort/reasoning-effort parameter, deciding whether to upgrade/downgrade after benchmarking, or verifying a model name is still current
references/cache-and-cost.mdAn agent's call cadence vs. its provider's prompt-cache TTL, ranking candidate models by cost tier, applying a cost ceiling inherited from a caller, or picking genuinely independent models/providers for adversarial or second-opinion review
references/harness-mapping.mdWriting the actual model:-equivalent field for GitHub Copilot CLI, Claude Code, or VS Code custom agents, understanding each harness's resolution order and fallback-list support, or extending this skill's guidance to a harness not yet covered

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

コメント

ログインなしで書き込めます。使ってみた感想、質問、動かなかった環境などを気軽にどうぞ。

まだコメントはありません。最初のコメントを書いてみませんか。

送信するとすぐに公開されます。メールアドレスや電話番号などの個人情報は書かないでください。URL は 2 つまでです。

0 / 1000 文字

このフォームは reCAPTCHA で保護されており、Google のプライバシーポリシーと利用規約が適用されます。

同じリポジトリのスキル

3d-rotation-theory

無料

Abstract math for 3D rotation representations and conversions. USE FOR: understanding axis-angle representation; Euler angles (yaw/pitch/roll, convention choices); rotation matrices (SO(3)); quaternion-based rotation; converting between representations; identifying and avoiding gimbal lock; composing rotations; half-angle formula; right-hand rule. DO NOT USE FOR: quaternion arithmetic details (use quaternion-algebra); slerp/lerp interpolation (use interpolation-on-manifolds); coordinate system definitions (use coordinate-system-conversions).

日本語の概要は準備中です。原文の説明を表示しています。

NetFabric/NetFabric.Numerics362026年8月10日 更新

agent-graph-orchestration

無料

Design and scaffold multi-agent squads/swarms using the graph orchestration pattern (nodes = agents, edges = dependencies/data flow). USE FOR: designing a coordination topology for a multi-agent squad (sequential pipeline, parallel fan-out/aggregation, branching, feedback loop, cyclic); building an orchestrator agent that delegates to specialist agents; setting up an adversarial/tribunal code or document review with independent reviewers; scaffolding a squad's agent files for a target coding-agent harness (e.g. Copilot CLI or Claude Code custom agents). DO NOT USE FOR: authoring a single agent's frontmatter/prompt body (use copilot-cli-custom-agents, claude-code-custom-agents, or the target harness's own skill); SKILL.md authoring (use create-skill); ad-hoc multi-agent chat with no explicit topology.

日本語の概要は準備中です。原文の説明を表示しています。

NetFabric/NetFabric.Numerics362026年8月10日 更新

agent-harness-logs

無料

Find, inspect, query, redact, and troubleshoot logs, transcripts, sessions, and OpenTelemetry data from coding-agent harnesses on Windows, macOS, Linux, and WSL. USE FOR: GitHub Copilot in VS Code; Copilot CLI; Claude Code; OpenAI Codex; Gemini CLI; OpenCode; Cursor, Windsurf, Kiro, or other VS Code-family editors; locating log files; diagnosing tool or model failures; correlating turns, subagents, and tool calls; querying JSONL, plaintext, or SQLite without spending model tokens; preparing sanitized diagnostics. DO NOT USE FOR: application logs unrelated to coding agents; uploading raw logs; treating unstable transcript schemas as public APIs.

日本語の概要は準備中です。原文の説明を表示しています。

NetFabric/NetFabric.Numerics362026年8月10日 更新

agents-md

無料

Author and maintain AGENTS.md files — the open, tool-agnostic format for giving coding agents project context. Use when: creating AGENTS.md; writing an agent instructions file; root or nested AGENTS.md for monorepos; migrating CLAUDE.md, AGENT.md, .cursor/rules, or copilot-instructions.md to AGENTS.md; greenfield project setup before code exists; brownfield onboarding for an existing codebase; documenting build/test/lint commands, code style, PR/commit conventions, and security considerations for agents; AGENTS.md discovery and precedence rules. Works for any language, framework, or project size. DO NOT USE FOR: authoring SKILL.md skill files (use create-skill); APM package/plugin authoring (use apm-producer); language-specific style rules (use language skills e.g. csharp-best-practices).

日本語の概要は準備中です。原文の説明を表示しています。

NetFabric/NetFabric.Numerics362026年8月10日 更新

angles-and-circular-arithmetic

無料

Abstract math for angle measurement, unit systems, and circular arithmetic. USE FOR: implementing angle types; converting between degrees/radians/gradians/revolutions; reducing angles to canonical range [0, Full); computing reference angles and quadrants; implementing safe modular arithmetic with negative inputs; angle addition and subtraction; classifying angles by sign and size. DO NOT USE FOR: trigonometric functions (use trigonometric-functions); coordinate conversions (use coordinate-system-conversions); interpolation (use interpolation-on-manifolds).

日本語の概要は準備中です。原文の説明を表示しています。

NetFabric/NetFabric.Numerics362026年8月10日 更新

claude-code-custom-agents

無料

Create and maintain Claude Code custom subagents (.claude/agents/*.md files). USE FOR: authoring project (.claude/agents/) or user (~/.claude/agents/) subagent files; frontmatter fields (name, description, tools, disallowedTools, model, permissionMode, maxTurns, skills, mcpServers, hooks, memory, background, effort, isolation, color, initialPrompt); writing descriptions that drive automatic delegation and 'use proactively' phrasing; invoking subagents via natural language, @-mention, session-wide --agent, or the Agent tool; restricting which subagent types a coordinator can spawn with Agent(name1, name2); nested subagents and depth limits; built-in subagents (Explore, Plan, general-purpose, claude); troubleshooting a subagent that won't load or isn't invoked. DO NOT USE FOR: GitHub Copilot CLI custom agents (use copilot-cli-custom-agents); agent teams/teammates (a different, session-spanning feature); MCP server authoring; SKILL.md authoring (use create-skill).

日本語の概要は準備中です。原文の説明を表示しています。

NetFabric/NetFabric.Numerics362026年8月10日 更新

NetFabric のスキルをすべて見る

このスキルの問題を報告する