skills.rest
wecode-ai
wecode-aiOFFICIAL
Explore all skills by this author →
GitHub

wegent-knowledge

Manage and search Wegent knowledge bases and documents using RAG retrieval.

772|135|Updated Jan 10, 2026
Search Tags:#mcp#rAG#Semantic Search#Indexing#Knowledge Base#Document Management#wegent
One-click install
npx skills add https://github.com/wecode-ai/Wegent --skill wegent-knowledge
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: wegent-knowledge
Source: https://github.com/wecode-ai/Wegent/tree/main/backend/init_data/skills/wegent-knowledge
Command: npx skills add https://github.com/wecode-ai/Wegent --skill wegent-knowledge

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It gives AI agents programmatic access to Wegent knowledge bases, eliminating manual switching between the chat interface and the knowledge management UI to list, create, update, read, and semantically search documents.

Core Features & Use Cases

  • Knowledge Base Management: List knowledge bases across personal, group, and organization scopes, and create new notebook or classic knowledge bases.
  • Document Operations: Create documents from text, base64 files, or web URLs, read long documents incrementally with offset pagination, and update plain-text document content with automatic re-indexing.
  • RAG Search: Run semantic retrieval over a knowledge base with query hints, folder scoping, and document ID filters.
  • Use Case: Ask the agent to search your team's knowledge base for deployment steps, then create a new meeting-notes document summarizing the findings, with indexing triggered automatically.

Quick Start

List all my accessible knowledge bases and then search the engineering one for documents about deployment configuration.

Frequently Asked Questions about wegent-knowledge

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

FAQPage Schema
How do I search a knowledge base with RAG retrieval?▼

Call wegent_kb_search_knowledge_base with the knowledge_base_id and a query string. You can narrow results with document_ids or folder_ids, pass semantic hints via search_hints, and cap output with max_results up to 50.

How to create a document in a Wegent knowledge base?▼

Use wegent_kb_create_document with the target knowledge_base_id, a name, and a source_type of text, file, or web. Text takes inline content, file takes base64 data with an extension, and web fetches a URL, with indexing and summary triggered by default.

Can I update PDF or Word documents in a knowledge base?▼

Inline content updates only work for text documents and plain-text files like txt, md, and markdown. Binary formats such as pdf or docx must be replaced by creating or re-uploading the source file instead of editing content directly.

Why does my new document show pending status?▼

Document indexing runs asynchronously after creation or update, so documents display a pending status until RAG indexing completes. Wait for indexing to finish before expecting the document to appear in search results.

How do I read long documents without truncation?▼

Read incrementally with wegent_kb_read_document_content using offset and limit parameters. Start with the backend default limit, then continue with offset equal to the previous offset plus returned length while has_more is true.

Related Skills

Explore More Software Engineering Skills→
atro2829-hubatro2829-hub

ASR

Transcribe spoken audio from files or base64 input into text.

Community
Intermediate
h4vzzh4vzz
33

context-retrieval

Retrieve and rank knowledge base passages with embedding search and reranking.

Community
Advanced
Soul-Brews-StudioSoul-Brews-Studio
116

vault

Connect Obsidian, Logseq, and Markdown folders to Oracle for cross-vault search.

Official
Intermediate
qodex-aiqodex-ai
39

rag-agent-builder

Build RAG applications with vector databases, embeddings, and retrieval strategies.

Official
Advanced