技能 优化
- 作者仓库星标 1,812
- 作者仓库 skills
When to use
Use this skill when you need to:
- Improve whether a skill is actually applied by models
- Diagnose why some criteria fail across all models
- Prevent a skill from making outputs worse
- Refactor skill text for stronger retrieval under context pressure
- Build repeatable benchmark loops and release gates
Optimization loop (default workflow)
- Measure baseline and skill-on behavior (per model, per scenario, per criterion)
- Find failure pattern:
- universal failure (0% with skill)
- model-specific weakness
- regression (negative delta)
- Edit for salience:
- add explicit triggers
- add concrete integrated examples
- tighten checklists and decision rules
- Re-run evals and compare deltas
- Ship with guardrails (documented gate + run history + follow-up issues)
How to use
Read individual rule files for detailed procedures and templates:
- rules/benchmark-loop.md - End-to-end benchmark loop and scoring
- rules/activation-design.md - Improve retrieval and instruction uptake
- rules/context-budget.md - Reduce token cost without losing behavior
- rules/regression-triage.md - Diagnose and fix skill-on regressions
- rules/release-gates.md - Go/no-go criteria before shipping skill updates
Practical heuristics
- Prefer few high-signal rules over many soft recommendations
- Put fragile, high-value behaviors in top-level checklists
- Include at least one integrated example per common scenario
- Add explicit wording for what must not be omitted
- Track gains/losses with with-skill vs without-skill comparisons
- 流狐分类
- AI 智能
- 作者声明 Agent
- 未找到明确声明；不据此推断已兼容或已测试
- 静态检查
- 88 / 100 · 启发式扫描，不代表运行安全
- 作者 / 版本 / 许可
- @mcollina · 未声明 license
- 流狐 Token 估算
- 低消耗
- 流狐接入估算
- 即装即用
- 是否需要外部 API Key
- 未发现要求
- 检测到的系统要求
- 未声明
- 底层运行要求
- 未声明
- 检测到的文件与系统行为
-
- 只读
- 允许写入 / 修改
- 检测到的网络行为
- 仅限本地
- 安装命令数
- 无（仅作为资料）
档案由构建时根据 SKILL.md 与安装命令自动衍生，可能与作者实际意图存在差异。
需要注意： 未限定 allowed-tools，默认拥有全部工具权限。
作者没有在当前 SKILL.md 中定义固定输出样例。 When to use
Use this skill when you need to: Improve whether a skill is actually applied by models Diagnose why some criteria fail across all models
Optimization loop (default workflow)
Measure baseline and skill-on behavior (per model, per scenario, per criterion) Find failure pattern: universal failure (0% with skill)
How to use
Read individual rule files for detailed procedures and templates: rules/benchmark-loop.md - End-to-end benchmark loop and scoring rules/activation-design.md - Improve retrieval and instruction uptake
Practical heuristics
Prefer few high-signal rules over many soft recommendations Put fragile, high-value behaviors in top-level checklists Include at least one integrated example per common scenario
## When to use
Use this skill when you need to:
- Improve whether a skill is actually applied by models
- Diagnose why some criteria fail across all models
- Prevent a skill from making outputs worse
- Refactor skill text for stronger retrieval under context pressure
- Build repeatable benchmark loops and release gates
## Optimization loop (default workflow)
1. **Measure baseline and skill-on behavior** (per model, per scenario, per criterion)
2. **Find failure pattern**:
- universal failure (0% with skill)
- model-specific weakness
- regression (negative delta)
3. **Edit for salience**:
- add explicit triggers
- add concrete integrated examples
- tighten checklists and decision rules
4. **Re-run evals** and compare deltas
5. **Ship with guardrails** (documented gate + run history + follow-up issues)
## How to use
Read individual rule files for detailed procedures and templates:
- [rules/benchmark-loop.md](rules/benchmark-loop.md) - End-to-end benchmark loop and scoring
- [rules/activation-design.md](rules/activation-design.md) - Improve retrieval and instruction uptake
- [rules/context-budget.md](rules/context-budget.md) - Reduce token cost without losing behavior
- [rules/regression-triage.md](rules/regression-triage.md) - Diagnose and fix skill-on regressions
- [rules/release-gates.md](rules/release-gates.md) - Go/no-go criteria before shipping skill updates
## Practical heuristics
- Prefer **few high-signal rules** over many soft recommendations
- Put fragile, high-value behaviors in **top-level checklists**
- Include at least one **integrated example** per common scenario
- Add explicit wording for what must **not** be omitted
- Track gains/losses with **with-skill vs without-skill** comparisons 证据边界与执行链路
作者原文负责流程事实；流狐只索引当前章节、要点、文件与命令。
章节 -> When to use → Optimization loop (default workflow) → How to use → Practical heuristics
要点 -> Measure baseline and skill-on behavior · Find failure pattern · Edit for salience · Re-run evals · Ship with guardrails · few high-signal rules · top-level checklists · integrated example
文件/命令 -> rules/benchmark-loop.md · rules/activation-design.md · rules/context-budget.md · rules/regression-triage.md · rules/release-gates.md · Go/no-go · gains/losses
内容 SHA-256 -> ce9b4b444e15
方法与流程
适用与边界
原文中的明确线索
rules/benchmark-loop.md、rules/activation-design.md、rules/context-budget.md、rules/regression-triage.md、rules/release-gates.md、Go/no-go、gains/losses