/create-skills-via-llms
Create high-quality Rust crate skills from llms.txt
> /plugin marketplace add zhanghandong/rust-skills > /plugin install rust-skills@rust-skills
How it fires
How this command gets triggered: by you, by Claude, or both.
- Fires itselfClaude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/create-skills-via-llms
Context preview
What this command does when you run it.
Create high-quality Rust crate skills from llms.txt
Command definition
create-skills-via-llms.mddescription: Create high-quality Rust crate skills from llms.txt
argument-hint: <crate_name> <llms_path> [version] [description]
Create high-quality skills for a Rust crate based on llms.txt documentation.
Arguments: $ARGUMENTS
- First argument: crate_name (required) - the Rust crate name (e.g., tokio, serde)
- Second argument: llms_path (required) - local path to the llms.txt file
- Third argument: version (optional) - the crate version (e.g., "1.40.0", "2.0.0")
- Fourth argument: description (optional) - additional requirements or information
---
Task: Create {crate_name} Skills
**llms.txt file location**: {llms_path}
---
Skill Quality Standards
Each skill must include the following structure:
SKILL.md Structure
---
name: {crate_name}-{feature}
description: |
CRITICAL: Use for {crate_name} {feature} questions. Triggers on:
{keyword1}, {keyword2}, {keyword3}, "{common question}",
{中文关键词1}, {中文关键词2}, {中文问题}
---
# {CrateName} {Feature} Skill
> **Version:** {crate_name} {version} | **Last Updated:** {YYYY-MM-DD}
>
> Check for updates: https://crates.io/crates/{crate_name}
You are an expert at the Rust `{crate_name}` crate. Help users by:
- **Writing code**: Generate Rust code following the patterns below
- **Answering questions**: Explain concepts, troubleshoot issues, reference documentation
## Documentation
Refer to the local files for detailed documentation:
- `./references/{file1}.md` - {description}
- `./references/{file2}.md` - {description}
## IMPORTANT: Documentation Completeness Check
**Before answering questions, Claude MUST:**
1. Read the relevant reference file(s) listed above
2. If file read fails or file is empty:
- Inform user: "本地文档不完整,建议运行 `/sync-crate-skills {crate_name} --force` 更新文档"
- Still answer based on SKILL.md patterns + built-in knowledge
3. If reference file exists, incorporate its content into the answer
## Key Patterns
{Core code patterns, 3-5 most commonly used patterns}
## API Reference Table
| Function/Type | Description | Example |
|---------------|-------------|---------|
| ... | ... | ... |
## Deprecated Patterns (Don't Use)
| Deprecated | Correct | Notes |
|------------|---------|-------|
| ... | ... | ... |
## When Writing Code
1. {Best practice 1}
2. {Best practice 2}
3. ...
## When Answering Questions
1. {Key point 1}
2. {Key point 2}
3. ...
````
### References Directory
Each skill's `references/` directory contains detailed documentation:
- API reference documentation
- Configuration options details
- Advanced usage examples
- Feature-specific configurations
---
## Instructions
### 1. Read llms.txt and Analyze
1. **Read the entire llms.txt** content
2. **Identify content domains**: Find functional modules that can be separate skills
3. **Analyze each domain**:
- What are the core concepts?
- What APIs/configuration options exist?
- What are common usage patterns?
- What content needs detailed documentation?
### 1.5 Confirm Version Number
If the user did not provide a version number (third argument):
1. Use the AskUserQuestion tool to ask the user for the current version
2. Version format examples: "1.40.0", "2.0.0", "latest"
3. Use the version number for all SKILL.md Version fields
### 2. Output Detailed Plan
Output to `~/tmp/{YYYYMMDDHHmm}-{crate_name}-skills-plan.md`:
````markdown
# {CrateName} Skills Plan
## Analysis Summary
- Crate: {crate_name}
- Version: {version}
- Main functional domains: ...
## Skill List
### 1. {crate_name}-{feature1}
**Trigger conditions**: "...", "...", "..."
**Core content**: ...
**Reference files**:
- {file1}.md - {description}
- {file2}.md - {description}
### 2. {crate_name}-{feature2}
...
````
### 3. Create Skills
For each skill:
1. **Create directory structure**:~/.claude/skills/{crate_name}-{feature}/ ├── SKILL.md └── references/ ├── {api-reference}.md └── {detailed-guide}.md
2. **Write SKILL.md**:
- Follow the quality standards above
- Keep SKILL.md concise (<200 lines)
- Put complex content in references/
3. **Write reference files**:
- Complete API reference
- Configuration options tables
- Detailed code examples
- Feature-specific content
### 4. Content Allocation Principles
| Content Type | Location |
|--------------|----------|
| Core patterns (3-5) | SKILL.md |
| Complete API reference | references/ |
| Configuration options details | references/ |
| Feature-specific config | references/ |
| Advanced usage/edge cases | references/ |
| Deprecated patterns table | SKILL.md |
| Best practices | SKILL.md |
---
## Quality Checklist
- [ ] Each SKILL.md has CSO-optimized description with "CRITICAL:" prefix
- [ ] Each SKILL.md description includes Chinese trigger keywords
- [ ] Each SKILL.md has version info and update date
- [ ] Each SKILL.md has "You are an expert..." role definition
- [ ] Each SKILL.md has Documentation navigation list
- [ ] Each SKILL.md has "Documentation Completeness Check" section
- [ ] Each SKILL.md has Key Patterns code examples
- [ ] Each SKILL.md has Deprecated Patterns table (if applicable)
- [ ] Each SKILL.md has "When Writing Code" best practices
- [ ] Each SKILL.md has "When Answering Questions" guidelines
- [ ] Complex content has been split into references/ directory
- [ ] Code examples use latest Rust idioms
- [ ] No redundant documentation files (README.md, etc.)
- [ ] Skills created directly in `~/.claude/skills/` for auto-discovery
---
## Output Location
All skills are created in: `~/.claude/skills/{crate_name}-*/`
This is the local dynamic skills directory, not committed to the rust-skills repository.Read more
description: Create high-quality Rust crate skills from llms.txt argument-hint: <crate_name> <llms_path> [version] [description]
Create high-quality skills for a Rust crate based on llms.txt documentation.
Arguments: $ARGUMENTS
- First argument: crate_name (required) - the Rust crate name (e.g., tokio, serde)
- Second argument: llms_path (required) - local path to the llms.txt file
- Third argument: version (optional) - the crate version (e.g., "1.40.0", "2.0.0")
- Fourth argument: description (optional) - additional requirements or information
---
Task: Create {crate_name} Skills
**llms.txt file location**: {llms_path}
---
Skill Quality Standards
Each skill must include the following structure:
SKILL.md Structure
---
name: {crate_name}-{feature}
description: |
CRITICAL: Use for {crate_name} {feature} questions. Triggers on:
{keyword1}, {keyword2}, {keyword3}, "{common question}",
{中文关键词1}, {中文关键词2}, {中文问题}
---
# {CrateName} {Feature} Skill
> **Version:** {crate_name} {version} | **Last Updated:** {YYYY-MM-DD}
>
> Check for updates: https://crates.io/crates/{crate_name}
You are an expert at the Rust `{crate_name}` crate. Help users by:
- **Writing code**: Generate Rust code following the patterns below
- **Answering questions**: Explain concepts, troubleshoot issues, reference documentation
## Documentation
Refer to the local files for detailed documentation:
- `./references/{file1}.md` - {description}
- `./references/{file2}.md` - {description}
## IMPORTANT: Documentation Completeness Check
**Before answering questions, Claude MUST:**
1. Read the relevant reference file(s) listed above
2. If file read fails or file is empty:
- Inform user: "本地文档不完整,建议运行 `/sync-crate-skills {crate_name} --force` 更新文档"
- Still answer based on SKILL.md patterns + built-in knowledge
3. If reference file exists, incorporate its content into the answer
## Key Patterns
{Core code patterns, 3-5 most commonly used patterns}
## API Reference Table
| Function/Type | Description | Example |
|---------------|-------------|---------|
| ... | ... | ... |
## Deprecated Patterns (Don't Use)
| Deprecated | Correct | Notes |
|------------|---------|-------|
| ... | ... | ... |
## When Writing Code
1. {Best practice 1}
2. {Best practice 2}
3. ...
## When Answering Questions
1. {Key point 1}
2. {Key point 2}
3. ...
````
### References Directory
Each skill's `references/` directory contains detailed documentation:
- API reference documentation
- Configuration options details
- Advanced usage examples
- Feature-specific configurations
---
## Instructions
### 1. Read llms.txt and Analyze
1. **Read the entire llms.txt** content
2. **Identify content domains**: Find functional modules that can be separate skills
3. **Analyze each domain**:
- What are the core concepts?
- What APIs/configuration options exist?
- What are common usage patterns?
- What content needs detailed documentation?
### 1.5 Confirm Version Number
If the user did not provide a version number (third argument):
1. Use the AskUserQuestion tool to ask the user for the current version
2. Version format examples: "1.40.0", "2.0.0", "latest"
3. Use the version number for all SKILL.md Version fields
### 2. Output Detailed Plan
Output to `~/tmp/{YYYYMMDDHHmm}-{crate_name}-skills-plan.md`:
````markdown
# {CrateName} Skills Plan
## Analysis Summary
- Crate: {crate_name}
- Version: {version}
- Main functional domains: ...
## Skill List
### 1. {crate_name}-{feature1}
**Trigger conditions**: "...", "...", "..."
**Core content**: ...
**Reference files**:
- {file1}.md - {description}
- {file2}.md - {description}
### 2. {crate_name}-{feature2}
...
````
### 3. Create Skills
For each skill:
1. **Create directory structure**:~/.claude/skills/{crate_name}-{feature}/ ├── SKILL.md └── references/ ├── {api-reference}.md └── {detailed-guide}.md
2. **Write SKILL.md**:
- Follow the quality standards above
- Keep SKILL.md concise (<200 lines)
- Put complex content in references/
3. **Write reference files**:
- Complete API reference
- Configuration options tables
- Detailed code examples
- Feature-specific content
### 4. Content Allocation Principles
| Content Type | Location |
|--------------|----------|
| Core patterns (3-5) | SKILL.md |
| Complete API reference | references/ |
| Configuration options details | references/ |
| Feature-specific config | references/ |
| Advanced usage/edge cases | references/ |
| Deprecated patterns table | SKILL.md |
| Best practices | SKILL.md |
---
## Quality Checklist
- [ ] Each SKILL.md has CSO-optimized description with "CRITICAL:" prefix
- [ ] Each SKILL.md description includes Chinese trigger keywords
- [ ] Each SKILL.md has version info and update date
- [ ] Each SKILL.md has "You are an expert..." role definition
- [ ] Each SKILL.md has Documentation navigation list
- [ ] Each SKILL.md has "Documentation Completeness Check" section
- [ ] Each SKILL.md has Key Patterns code examples
- [ ] Each SKILL.md has Deprecated Patterns table (if applicable)
- [ ] Each SKILL.md has "When Writing Code" best practices
- [ ] Each SKILL.md has "When Answering Questions" guidelines
- [ ] Complex content has been split into references/ directory
- [ ] Code examples use latest Rust idioms
- [ ] No redundant documentation files (README.md, etc.)
- [ ] Skills created directly in `~/.claude/skills/` for auto-discovery
---
## Output Location
All skills are created in: `~/.claude/skills/{crate_name}-*/`
This is the local dynamic skills directory, not committed to the rust-skills repository.AI-powered Rust development assistant with meta-cognition framework
Other commands on rust-skills.
- /achievement
View coding achievements, stats, and progress
Open command - /ai-daily
Generate AI daily/weekly news report from Reddit communities
Open command - /audit
Heavy-weight security and safety audit using os-checker tools.
Open command - /cache-clean
Clean Rust docs cache
Open command - /cache-status
Show Rust docs cache status
Open command - /clean-crate-skills
Remove local dynamic crate skills
Open command

