Claude-skill-registry knowledge-discovery
Discover relevant skills and knowledge using tiered approach. SKIP for simple tasks, QUICK for single-skill, STANDARD for multi-skill, DEEP for full feature implementation. Auto-selects tier based on task complexity.
install
source · Clone the upstream repo
git clone https://github.com/majiayu000/claude-skill-registry
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/knowledge-discovery" ~/.claude/skills/majiayu000-claude-skill-registry-knowledge-discovery && rm -rf "$T"
manifest:
skills/data/knowledge-discovery/SKILL.mdsource content
Knowledge Discovery
Adaptive discovery skill with tiered efficiency for different task complexities.
Discovery Tiers
| Tier | When to Use | Files Read | Time Cost |
|---|---|---|---|
| SKIP | Simple/obvious tasks, single file edits | 0 | ~0s |
| QUICK | Single-skill tasks, known patterns | Quick ref only | ~2s |
| STANDARD | Multi-skill tasks, unfamiliar patterns | INDEX + GRAPH | ~10s |
| DEEP | Full feature implementation, CRUD flows | All 4 steps | ~15-20s |
Tier Selection Guide
START: What's the task? │ ├─ "Fix typo" / "Add comment" / "Simple edit" │ └─ SKIP (no discovery needed) │ ├─ "Write tests" / "Add validation" / "Single pattern" │ └─ QUICK (use inline reference below) │ ├─ "Design API" / "Debug error" / "Multi-step task" │ └─ STANDARD (read SKILL-INDEX + CONTEXT-GRAPH) │ └─ "Implement CRUD" / "Add feature" / "Full workflow" └─ DEEP (full 4-step protocol)
QUICK Tier: Inline Reference
Use this for single-skill tasks without reading any files:
| Task | Skill | No Lookup Needed |
|---|---|---|
| Entity/DTO/AppService | | ✓ |
| DbContext/Migration | | ✓ |
| Input validation | | ✓ |
| Permissions/Auth | | ✓ |
| Unit/Integration tests | | ✓ |
| E2E tests | | ✓ |
| Query optimization | | ✓ |
| API design | | ✓ |
| Technical design doc | | ✓ |
| Debug/errors | | ✓ |
| Docker/.NET | | ✓ |
| Git advanced | | ✓ |
| Security audit | | ✓ |
Common Error → Skill:
| Error | Skill |
|---|---|
| N+1 query | |
| Authorization failed | |
| Validation failed | |
| DbUpdateException | |
| Task was canceled | |
STANDARD Tier Protocol
For multi-skill tasks, read these files:
1. DISCOVER → Check SKILL-INDEX.md for relevant skills 2. RELATE → Check CONTEXT-GRAPH.md for dependencies
DEEP Tier Protocol
For full feature implementation:
1. DISCOVER → Check SKILL-INDEX.md for relevant skills 2. RELATE → Check CONTEXT-GRAPH.md for dependencies 3. REFERENCE → Read /knowledge/ for shared patterns 4. FOLLOW → Use /flows/ for multi-step workflows
Discovery by Task Type
Creating a New Entity/Feature
- Check flow: flows/crud-implementation.md
- Primary skills:
,abp-framework-patterns
,efcore-patternsfluentvalidation-patterns - Knowledge:
Adding Validation
- Primary skill:
fluentvalidation-patterns - Knowledge:
Implementing Authorization
- Primary skill:
openiddict-authorization - Related:
,security-patternsabp-framework-patterns - Knowledge: knowledge/conventions/permissions.md
Writing Tests
- Primary skill:
xunit-testing-patterns - Knowledge:
Debugging an Error
- Check index: Search error message in SKILL-INDEX.md (Error Message section)
- Primary skill:
debugging-patterns - Related: Check skill by keyword match
Designing an API
- Primary skills:
,api-design-principlestechnical-design-patterns - Related:
,domain-modelingrequirements-engineering
Optimizing Queries
- Primary skill:
linq-optimization-patterns - Related:
efcore-patterns - Check index: Search "N+1" in SKILL-INDEX.md
Skill Layer Loading
Load skills in dependency order:
Layer 1 (Foundations) → Load first csharp-advanced-patterns, dotnet-async-patterns, error-handling-patterns Layer 2 (Framework) → Load second abp-framework-patterns, efcore-patterns, fluentvalidation-patterns Layer 3 (Features) → Load third xunit-testing-patterns, security-patterns, distributed-events Layer 4 (Workflows) → Load as needed feature-development-workflow
Index Files Reference
| File | Purpose | Location |
|---|---|---|
| SKILL-INDEX.md | Find skills by task/keyword/error | SKILL-INDEX.md |
| CONTEXT-GRAPH.md | Skill dependencies, layers, relationships | CONTEXT-GRAPH.md |
| knowledge/INDEX.md | Shared patterns and conventions | knowledge/INDEX.md |
| flows/INDEX.md | Multi-step workflows | flows/INDEX.md |
Quick Lookup Commands
By Task
"I need to create an entity" → abp-framework-patterns + efcore-patterns "I need to validate input" → fluentvalidation-patterns "I need to add permissions" → openiddict-authorization "I need to write tests" → xunit-testing-patterns "I need to optimize queries" → linq-optimization-patterns
By Error
"N+1 query" → linq-optimization-patterns "Authorization failed" → openiddict-authorization "Validation failed" → fluentvalidation-patterns "DbUpdateException" → efcore-patterns
By Keyword
"Entity" → abp-framework-patterns "DbContext" → efcore-patterns "FluentValidation" → fluentvalidation-patterns "[Authorize]" → openiddict-authorization "xUnit" → xunit-testing-patterns
Integration
This skill should be invoked:
- Automatically at the start of complex tasks
- When the user asks "what skills do I need for..."
- Before implementing multi-step workflows
- When encountering unfamiliar patterns or errors
Related Skills
After discovery, typically load:
- Core ABP patternsabp-framework-patterns
- Database layerefcore-patterns
- Input validationfluentvalidation-patterns
- Permissionsopeniddict-authorization
- Testingxunit-testing-patterns