skills.rest
contexthub-md
contexthub-mdOFFICIAL
Explore all skills by this author →
GitHub

discover

Searches a project's knowledge base using semantic search and RAG over indexed repos and docs.

Updated May 12, 2026
Search Tags:#semantic search#rag#knowledge base#mcp#code search#context discovery#contexthub
One-click install
npx skills add https://github.com/contexthub-md/marketplace --skill discover-contexthub-md
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: discover
Source: https://github.com/contexthub-md/marketplace/tree/main/contexthub/skills/discover
Command: npx skills add https://github.com/contexthub-md/marketplace --skill discover-contexthub-md

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Finding relevant code, documentation, or project context across a large knowledge base is slow when done manually. This Skill runs semantic search and RAG queries against your project's indexed repositories and documents so you get ranked snippets or direct answers to questions about how things work. ## Core Features & Use Cases - Semantic Search: Query indexed project content with ch_discover_context and get ranked snippets with sources. - RAG Answers: Set include_answer: true to receive an LLM-generated answer synthesized from retrieved content. - Provider & Scope Control: Search the web via Tavily, library docs via Context7, filter by content type (code, docs, visual), or scope to a specific project. - Use Case: Ask "how does authentication work?" and receive either the most relevant code/doc snippets or a synthesized answer drawn from your project's knowledge base. ## Quick Start Ask the assistant to use the discover skill to search the project knowledge base for how authentication works.

Frequently Asked Questions about discover

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search my project's knowledge base with semantic search?▼

Call ch_discover_context with your query to run semantic search over indexed repos and docs. Results return ranked content snippets with sources, and you can scope the search to a specific project using project_id.

How do I get a direct answer instead of search snippets?▼

Set include_answer to true when calling ch_discover_context to enable RAG mode. This returns an LLM-generated answer synthesized from the retrieved content alongside the source snippets.

When should I use discover vs kb vs research?▼

Use discover for general semantic search over your project's knowledge base. Use kb for specific indexed reference/library repos, research for multi-source queries including external library docs and the web, and ch_search_memory to recall past decisions.

Can I search the web or external library documentation?▼

Yes, set the provider parameter to tavily for web search or context7 for library documentation. You can also filter results by content_type such as code, docs, or visual.

What follow-up tools complement a discovery search?▼

Use ch_get_project for broader project context, ch_get_compass for a project health overview, ch_list_missions to see work in progress, and ch_search_memory for past decisions related to your query.

Related Skills

Explore More Software Engineering Skills→
spikelabspikelab

qmd-search

Searches a qmd-indexed resources knowledge base using semantic and keyword queries.

Community
Advanced
Prompt-or-Die-LabsPrompt-or-Die-Labs
7

create-memory

Create structured markdown memory files recording project learnings and decisions.

Official
Basic
MSApps-MobileMSApps-Mobile
8

knowledge-agent

Retrieve past decisions and insights from project memory with citations.

Official
Advanced
thoreinsteinthoreinstein
5

search

Runs parallel searches across codebases, documentation, and the web for references and patterns.

Community
Advanced