generate-skill
Skill7 starsChanged 6 months agoMITView on GitHub
引導使用者建立、更新、審查或最佳化高品質 Agent Skill 定義檔 (SKILL.md),包含官方規格、description 觸發語、references/scripts/assets/evals 結構,以及 5 種核心設計模式。
2 other projects use a version of this.
Adopted by
Most recently taken up by
- BryanBerger98/claude-skillstheir own version, first seen Jul 17, 2026
- AntJanus/skillboxtheir own version, first seen Jun 26, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- AntJanus/skillbox★ 2same name
- BryanBerger98/claude-skills★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository28 other files it gives its agents.
- AGENTS.mdAGENTS.md
- neo-azure-pipelinesSkill
- neo-clarificationSkill
- neo-clean-architectureSkill
- neo-code-reviewSkill
- neo-csharp-interface-generatorSkill
- neo-csharpSkill
- neo-dotnet-ef-coreSkill
- neo-dotnet-minimal-apisSkill
- neo-dotnet-mvcSkill
- neo-dotnetSkill
- neo-dotnet-tag-helperSkill
- neo-dotnet-webapiSkill
- neo-harness-projectSkill
- neo-harnessSkill
- neo-javascriptSkill
- neo-opentelemetrySkill
- neo-prSkill
- neo-python-managerSkill
- neo-pythonSkill
- neo-rabbitmqSkill
- neo-rustSkill
- neo-stop-slopSkill
- neo-sub-agentSkill
- neo-swiftSkill
- neo-swift-uiSkill
- neo-typescriptSkill
- neo-vueSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 2 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 11
If high-impact information is missing, ask concise follow-up questions before generating final files.
- Very specific: line 38
Ensure name is valid and matches the skill directory.
- Distinctive: line 39
Keep the main body procedural, concise, and grounded in provided expertise.
- Very specific: line 40
Use relative paths from the skill root for all references/, scripts/, and assets/.
- Very specific: line 42
Quality Gate: Before finalizing, check the output against .agents/skills/generate-skill/reference/AgentSkillsOfficialReference.md, …
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Skill Generator Specification
Perceive
- Receive the user's request to create, update, convert, review, or optimize an Agent Skill.
- Identify the target task, usage scenarios, pain points, expected outputs, target agent client, and whether the user already has source material such as docs, runbooks, examples, code, prompts, or previous failures.
- If high-impact information is missing, ask concise follow-up questions before generating final files. Do not invent domain rules that should come from the user or existing project material.
Reason
- Load Official Agent Skills Reference: Load and analyze
.agents/skills/generate-skill/reference/AgentSkillsOfficialReference.mdfirst. Treat it as the source of truth forSKILL.mdfrontmatter, directory structure, progressive disclosure, description triggering, scripts, references, assets, evals, and client implementation considerations. - Load Reference Patterns: Load and analyze
.agents/skills/generate-skill/reference/SkillCategory.mdto understand 5 core design patterns:- Tool Wrapper: Suitable for encapsulating APIs, libraries, or team development conventions.
- Generator: Suitable for consistently producing structured documents in specific formats (e.g., reports, templates).
- Reviewer: Suitable for systematically checking the quality of code or outputs (e.g., security, style).
- Inversion: Suitable for guided requirement gathering, ensuring the Agent does not guess answers prematurely.
- Pipeline: Suitable for complex tasks that require a strict execution order and cannot skip steps.
- Load Reference Templates: Load and analyze the standard templates under
.agents/skills/generate-skill/templates/(includingSKILL.md,evals.json,eval_queries.json, andpep723.py) to provide standard, high-quality boilerplate content to the user. - Load ReAct Guidelines: Load and analyze
.agents/skills/generate-skill/reference/ReAct_vs_CLI.mdwhen the skill involves local exploration, CLI behavior, autonomous debugging, cross-file analysis, CI/CD assistance, or another agentic operating-system workflow. - Requirement Matching: Based on the user's requirement, choose the most suitable pattern or combination of patterns. Use official Agent Skills format as the baseline; Perceive/Reason/Act is optional structure, not a mandatory official requirement.
- Progressive Disclosure Planning: Decide what belongs in
SKILL.mdversusreferences/,scripts/,assets/, andevals/. KeepSKILL.mdconcise and load detailed material only when needed. - Description & Validation Planning: Draft a trigger-focused
description, identify likely should-trigger and should-not-trigger prompts, and plan output quality checks for important or shareable skills.
Act
- Pattern Recommendation & Explanation: Present the recommended pattern(s), why they fit, and whether a ReAct-style autonomous workflow or simpler CLI/tool-wrapper approach is appropriate.
- Gather Necessary Variables: Guide the user to provide key information needed to generate or update the skill, including:
- Skill name (
name) - Trigger-focused description (
description) - Target agent/client and compatibility constraints
- Source material, domain rules, examples, and gotchas
- Required references, scripts, assets, templates, or evals
- Optional metadata, license, and allowed tool expectations
- Skill name (
- Generate File Content:
- Generate a standards-compliant
SKILL.mdwhose first line is---. - Ensure
nameis valid and matches the skill directory. - Keep the main body procedural, concise, and grounded in provided expertise.
- Use relative paths from the skill root for all
references/,scripts/, andassets/. - Guide the user to adopt the standard structure by generating supporting files (
evals/evals.json,eval_queries.json, and non-interactive scripts) based on the loaded templates.
- Generate a standards-compliant
- Quality Gate: Before finalizing, check the output against
.agents/skills/generate-skill/reference/AgentSkillsOfficialReference.md, especially frontmatter validity, description trigger quality, progressive disclosure, resource paths, script safety, and eval recommendations. - Provide Advanced Suggestions: Recommend trigger evals, output quality evals, reference files, templates, or scripts that would make the skill more reliable in real use.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

