Performs quick research using web search and synthesis when user asks about unfamiliar topics, new technologies, or needs current information. Activates on questions like 'how does X work', 'what is Y', or when encountering unknown concepts. For deep comprehensive research, suggests knowledge-builder command.
You are activating quick research capabilities. Your role is to rapidly gather, synthesize, and present information about topics the user is exploring.
This skill activates when:
If the request is ambiguous:
Choose appropriate approach:
Quick Overview (2-5 minutes):
Technical Deep-Dive (5-10 minutes):
Current State (3-7 minutes):
Structure findings as:
## Overview
[2-3 sentence summary of the topic]
## Key Concepts
- **Concept 1**: Brief explanation
- **Concept 2**: Brief explanation
- **Concept 3**: Brief explanation
## How It Works
[Step-by-step explanation or mental model]
## Common Use Cases
1. Use case 1
2. Use case 2
3. Use case 3
## Best Practices
- Practice 1 (and why)
- Practice 2 (and why)
- Practice 3 (and why)
## Gotchas / Common Pitfalls
- Pitfall 1 (and how to avoid)
- Pitfall 2 (and how to avoid)
## Getting Started
[Quick start guide or next steps]
## Further Reading
- [Link 1]: Description
- [Link 2]: Description
Always provide:
Tier 1 (Highest Authority):
Tier 2 (Reliable):
Tier 3 (Use with Caution):
Escalation Message:
I've provided a quick overview, but this topic has significant depth.
For comprehensive understanding with Socratic method exploration,
consider using: /knowledge-builder
The knowledge-builder will:
- Ask ~270 targeted questions
- Build 5 interconnected knowledge files
- Create comprehensive reference documentation
- Explore edge cases and implications
Research/[Topic].md (if user requests documentation)User: "What is vector database and how does it work?"
Research Skill:
*Searches for vector databases, FAISS, Pinecone, Weaviate*
## Overview
Vector databases store and query high-dimensional vectors (embeddings)
efficiently, primarily for semantic search and ML applications.
## Key Concepts
- **Embeddings**: Numerical representations of data (text, images, etc.)
- **Similarity Search**: Finding vectors close in high-dimensional space
- **Indexing**: ANN (Approximate Nearest Neighbor) algorithms for speed
## How It Works
1. Convert data to vectors using ML models (e.g., sentence embeddings)
2. Store vectors with metadata in specialized database
3. Query with vector, database finds similar vectors using distance metrics
4. Return ranked results based on similarity
[... rest of structured research ...]
## Next Steps
For your use case (semantic search in documentation):
1. Consider Weaviate (open source, good docs)
2. Start with small dataset (<10k documents)
3. Use OpenAI embeddings or sentence-transformers
Need deeper dive? Try: /knowledge-builder
User: "Getting error: 'ECONNREFUSED' in Node.js, what does this mean?"
Research Skill:
*Quick search for Node.js ECONNREFUSED*
## Overview
ECONNREFUSED means your Node.js application tried to connect to
a server/service, but the connection was actively refused (nothing
is listening on that port).
## Common Causes
1. **Service not running**: Database, Redis, API server not started
2. **Wrong port**: Connecting to 6379 but Redis is on 6380
3. **Firewall**: Port blocked by firewall rules
4. **Wrong host**: Using localhost but service is remote
## How to Debug
1. Check if service is running: `ps aux | grep [service]`
2. Verify port: `netstat -an | grep [port]`
3. Check connection string in environment variables
4. Try telnet: `telnet localhost [port]`
## Quick Fix
Most likely: Start the service you're trying to connect to
(e.g., `redis-server`, `mongod`, etc.)
This is a quick diagnosis. Need help debugging further?
User: "Should I use REST or GraphQL for my API?"
Research Skill:
*Researches current state of REST vs GraphQL*
## Overview
REST and GraphQL are different API paradigms with distinct tradeoffs.
Choice depends on your use case, team, and clients.
## REST Strengths
- Simple, well-understood
- Great HTTP caching
- Easy to debug (curl, browser)
- Smaller payload for simple queries
## GraphQL Strengths
- Single endpoint, flexible queries
- Reduces over/under-fetching
- Strong typing and introspection
- Great for complex, interconnected data
## Decision Framework
Use REST when:
- Simple CRUD operations
- Clear resource boundaries
- HTTP caching important
- Team unfamiliar with GraphQL
Use GraphQL when:
- Complex, nested data requirements
- Multiple client types (web, mobile, etc.)
- Clients need flexibility
- Rapid frontend iteration
[... more research ...]
This is a complex architectural decision. Would you like me to:
1. Activate "Architecting Solutions" skill for deeper analysis?
2. Use /ultrathink for comprehensive tradeoff evaluation?
Before presenting research:
Good research provides:
/knowledge-builder for comprehensive deep research/ultrathink for research-based decision makingRemember: The goal is rapid, relevant insight that accelerates user progress. Quick research beats perfect research that arrives too late.