-
-
Notifications
You must be signed in to change notification settings - Fork 524
Open
Labels
aiAI/ML powered featuresAI/ML powered featurescategory:communityCommunity engagement and outreachCommunity engagement and outreachenhancementNew feature or requestNew feature or requestmcproadmap:H1.3Roadmap task H1.3: Create example project folderRoadmap task H1.3: Create example project folderworkflowEnd-to-end workflow automationEnd-to-end workflow automation
Description
Category: 🌐 Community & Sharing - Config Sharing
Time: 4-6 hours
Add intelligent skill detection that automatically identifies missing skills and offers to install them.
Purpose
Enable Claude to detect when it needs a skill and automatically suggest or install it.
User Experience Examples
Auto-Detection:
User: "How do I create a React component with hooks?"
Claude: → Detects React → "I found a React skill. Install it? (30 sec)"
User: "Yes"
Claude: → install_skill(framework='react') → "React skill ready!"
Proactive Suggestion:
User: "Building a Django API"
Claude: → Found django.json, fastapi.json → "Install Django and FastAPI skills? (1 min)"
Features
- Topic Extraction: Parse user queries for framework/library names
- Skill Gap Analysis: Compare detected topics with installed skills
- API Search: Query api.skillseekersweb.com for matches
- Smart Suggestions: Single/multiple matches, confidence scoring
- Installation Modes: Ask first (default), Auto-install, Suggest only, Manual
MCP Tool: detect_and_suggest_skills
Parameters:
query: User's original queryauto_install: Auto-install without asking (default: false)max_suggestions: Max skills to suggest (default: 3)install_mode: 'ask', 'auto', 'suggest', 'manual'
Detection Algorithm
- Extract topics from query (regex + NLP keywords)
- Search API: exact match, category match, tag match
- Score matches (exact=10pts, category=5pts, tag=3pts)
- Check already installed skills
- Present top N options
Dependencies
- A1.1 (API endpoint) ✅ Complete
- A1.2 (MCP fetch_config) ✅ Complete
- A1.7 (install_skill) - Required for workflow
See: FLEXIBLE_ROADMAP.md - Task A1.8
Metadata
Metadata
Assignees
Labels
aiAI/ML powered featuresAI/ML powered featurescategory:communityCommunity engagement and outreachCommunity engagement and outreachenhancementNew feature or requestNew feature or requestmcproadmap:H1.3Roadmap task H1.3: Create example project folderRoadmap task H1.3: Create example project folderworkflowEnd-to-end workflow automationEnd-to-end workflow automation