minimal-implementation

以最小正确改动完成代码、Skill、规则、脚本、修复、重构和工作流修改，并提供可复核的验证证据。用于修改代码、Skill、规则文件、脚本、辅助函数或工具，或用户要求优化、改进、重构、稳定化和增强鲁棒性的场景，避免不必要的辅助代码、降级路径、抽象、依赖、整文件重写和无关重构。

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

78%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A concise, well-organized instruction skill with a clear decision procedure and explicit validation guidance. Its weakest area is actionability, where directives remain principled rather than concrete, and the validation step is a checkpoint rather than a full feedback loop.

Suggestions

Add one or two concrete worked examples of the decision ladder in action (e.g., a tiny before/after showing 'reuse existing helper' beating 'new abstraction') to lift actionability.

Turn the 验证 section into an explicit feedback loop: 'validate → if it fails, fix root cause → re-run the same check' so the workflow clarity reaches the highest anchor.

Make the 变更预算 step concrete by showing the minimal artifact to write down (e.g., a 3-line 'impact / acceptance / not-touched' template) instead of only describing it.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence — sections like '决策阶梯' and '安全边界' are dense directives with no padding or explanations of basic concepts; every token earns its place, matching the 'lean and efficient' anchor.	5 / 5
Actionability	It provides a concrete decision procedure (a 7-step ladder with '在第一个满足请求的层级停止') and a concrete output contract, but guidance stays at the principle level ('保留现有命名、结构和风格') with no executable examples or specific commands; not a 4 because, even for an instruction-only skill, several directives lack the concrete specificity that would make them directly actionable.	3 / 5
Workflow Clarity	The decision ladder is a clearly sequenced process with a stop condition, and the '验证' section supplies a validation checkpoint ('不要把 exit code 或旧产物单独当作成功证据…优先使用新鲜的定向测试、validator、哈希…'); not a 5 because there is no explicit validate→fix→retry feedback loop, only an emphasis on what counts as evidence.	4 / 5
Progressive Disclosure	The skill is under 50 lines, single-purpose, and organized into clear headers (核心规则, 决策阶梯, 变更预算, 安全边界, 输出契约, 验证) with no need for external references, matching the simple-skill exception that allows a 5 for well-organized sections.	5 / 5
	Total	17 / 20 Passed

Description

85%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, well-structured third-person description that clearly states both capability and trigger conditions with good keyword coverage. Its main weakness is distinctiveness: the optimization/refactor trigger terms are broad and risk overlapping with general editing skills.

Suggestions

Sharpen distinctiveness by leading with the signature behavior (e.g., 'Makes the smallest correct change that satisfies the request and produces verifiable evidence') so it is less likely to fire for ordinary refactors.

Add a disambiguating clause contrasting it with full-rewrite or architecture-design skills, since '重构/优化' currently overlaps heavily with general coding skills.

Tighten the trigger list toward the minimal-change intent (e.g., 'use when a change should stay surgical') to reduce conflict with broad optimization skills.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions across a broad surface — '完成代码、Skill、规则、脚本、修复、重构和工作流修改' plus '提供可复核的验证证据' — matching the comprehensive-coverage anchor; it is not the level below because it goes well beyond naming only the domain with 1-2 actions.	5 / 5
Completeness	Explicitly answers both 'what' ('以最小正确改动完成…修改，并提供可复核的验证证据') and 'when' ('用于…场景，或用户要求…的场景') with concrete trigger phrases; it is clearly not the level below where 'when' is only weakly implied.	5 / 5
Trigger Term Quality	Good natural-term coverage with synonyms ('优化、改进、重构、稳定化和增强鲁棒性') and concrete object nouns ('代码、Skill、规则文件、脚本、辅助函数或工具'); not a 5 because no file extensions or further common phrasings are included, leaving a few natural variations missing.	4 / 5
Distinctiveness Conflict Risk	The minimal-change philosophy is a recognizable niche, but trigger terms like '优化/重构/改进' are broad and would overlap with general coding and refactoring skills; it is not a 4 because the overlap risk with closely related edit/refactor skills is more than minor, and not a 2 because the 'minimal correct change' framing does carve out some distinct ground.	3 / 5
	Total	17 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: ooooooooooooooooooop/agent-tools
Path: skills/minimal-implementation/SKILL.md
Commit: 3677cfc

Reviewed: ~6 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.