knowledge_base

Agent Skills › siddsachar/row-bot › knowledge_base

knowledge_base

GitHub

管理用户个人知识库，整合知识图谱、文档存储与Wiki仓库。支持实体保存、关系链接、混合搜索及Markdown导出，实现结构化知识与非结构化文档的统一管理与检索。

bundled_skills/knowledge_base/SKILL.md siddsachar/row-bot

Trigger Scenarios

查询特定实体或主题的知识导入或整理大量记忆信息导出对话或知识到Wiki 查看知识库状态统计

Install

npx skills add siddsachar/row-bot --skill knowledge_base -g -y

More Options

Non-standard path

npx skills add https://github.com/siddsachar/row-bot/tree/main/bundled_skills/knowledge_base -g -y

Use without installing

npx skills use siddsachar/row-bot@knowledge_base

指定 Agent (Claude Code)

npx skills add siddsachar/row-bot --skill knowledge_base -a claude-code -g -y

安装 repo 全部 skill

npx skills add siddsachar/row-bot --all -g -y

预览 repo 内 skill

npx skills add siddsachar/row-bot --list

SKILL.md

Frontmatter

{
    "icon": "📚",
    "name": "knowledge_base",
    "tags": [
        "memory",
        "knowledge",
        "wiki"
    ],
    "author": "Row-Bot",
    "version": "1.0",
    "activation": {
        "phrases": [
            "knowledge base",
            "save to memory",
            "save this",
            "find it later",
            "remember this",
            "organize documents",
            "wiki vault",
            "knowledge graph"
        ],
        "examples": [
            "Organize these documents into my knowledge base"
        ],
        "keywords": [
            "knowledge",
            "memory",
            "save",
            "remember",
            "documents",
            "wiki",
            "vault",
            "graph"
        ],
        "negative_phrases": [
            "stale memory",
            "human tone"
        ]
    },
    "description": "Manage the user's personal knowledge base — knowledge graph, documents, and wiki vault.",
    "display_name": "Knowledge Base",
    "enabled_by_default": true
}

You are managing the user's personal knowledge base — a unified system combining a knowledge graph (entities + relations) with an Obsidian-compatible wiki vault of interconnected markdown files.

Your Knowledge System

The user's knowledge exists in three layers that you manage together:

Knowledge Graph — Entities (people, facts, preferences, events, places, projects) and typed relations stored in a searchable graph database.
Document Store — Uploaded files (PDF, DOCX, TXT) chunked and vectorized for semantic retrieval.
Wiki Vault — Auto-exported markdown files that mirror the knowledge graph. Each entity with sufficient content gets its own .md file with YAML frontmatter, [[wiki-links]] to related entities, and a connections section.

When to Use Wiki Tools

search_memory — Hybrid search across the knowledge graph (semantic + keyword
- graph expansion). Use when the user asks "what do you know about X", "search my knowledge base for Y", or when you need broader context than auto-recall provides.
wiki_read — When you need the full wiki article for a specific entity, including its frontmatter, connections, and complete description.
wiki_rebuild — After bulk memory operations (importing many facts, cleaning up duplicates) to regenerate all markdown files and indexes.
wiki_export_conversation — When the user says "save this conversation", "export our chat", or "add this to my wiki".
wiki_stats — When the user asks about the state of their knowledge base.

Building a Rich Knowledge Base

When the user shares information:

Save the entity with save_memory (appropriate category, descriptive content)
Link related entities with link_memories (use specific relation types)
The wiki vault updates automatically — no extra step needed

Best Practices

Prefer specific relation types: mother_of over related_to
Add aliases when entities have multiple names: "Mom, Sarah, Mother"
Use tags for cross-cutting themes: "family,birthday,march"
Rich descriptions export as better wiki articles
When the user mentions Obsidian, explain that the vault syncs automatically and they can open it in Obsidian to see backlinks, graph view, and search