skills.rest
mav-rik
mav-rikCOMMUNITY
Explore all skills by this author →
𝕏@MAVrik7GitHub

kb

Maintain a persistent markdown wiki with hybrid semantic and keyword search.

12|2|Updated May 11, 2026
Search Tags:#markdown#cli#wiki#knowledge-management#ai-agents#semantic-search#persistent-memory
One-click install
npx skills add https://github.com/mav-rik/kb-cli --skill kb-mav-rik
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: kb
Source: https://github.com/mav-rik/kb-cli/tree/main/content
Command: npx skills add https://github.com/mav-rik/kb-cli --skill kb-mav-rik

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

AI agents lose all context and stored knowledge when a conversation ends, forcing users to repeatedly re-explain background information, project details, and past decisions every time they interact with an agent. This Skill solves that problem by providing a persistent, searchable wiki that agents can use to store and retrieve knowledge across unlimited conversations, building a compounding knowledge base over time.

Core Features & Use Cases

  • Persistent Interlinked Markdown Wiki: Store knowledge as atomic, interconnected notes that future agents can traverse via pre-computed cross-references, making information discoverable both through search and link crawling.
  • Hybrid Semantic + Keyword Search: Find information using exact keyword matches for identifiers and error strings, or semantic similarity for conceptual questions, with results ranked via Reciprocal Rank Fusion for optimal relevance.
  • Automated Knowledge Workflows: Built-in guided workflows for ingesting new information, updating existing documents, and maintaining wiki health, including conflict resolution, linting, and schema management.
  • Use Case: A software team can use this Skill to store project documentation, API specifications, and past troubleshooting steps, so any AI agent working on the project has instant access to all context without requiring users to re-explain details in every new conversation.

Quick Start

Use the kb skill to store the project's new API authentication requirements in the wiki so they are available for all future agent conversations.

Frequently Asked Questions about kb

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

FAQPage Schema
How do I maintain persistent memory for AI agents across multiple conversations?▼

Persistent memory for AI agents is maintained by storing project context and decisions in a local interlinked markdown wiki. Agents retrieve this knowledge via semantic and keyword search, eliminating the need to re-explain background details in new conversations.

What's the best way to store project documentation so AI agents can search it semantically?▼

Storing project documentation in a SQLite-backed interlinked markdown wiki enables hybrid semantic and keyword search. Results are ranked using Reciprocal Rank Fusion, allowing agents to find exact error strings or conceptually related information efficiently.

How does hybrid search work in a local knowledge management wiki?▼

Hybrid search in a local knowledge management wiki combines exact keyword matching for identifiers with semantic similarity for conceptual queries. It ranks results via Reciprocal Rank Fusion, ensuring optimal relevance for both precise and abstract information retrieval.

Do I need the kb CLI to ingest and update markdown notes for cross-session memory?▼

Yes, the kb CLI is required to execute ingest, search, update, and lint operations on the wiki. It manages the SQLite backend and automated workflows, ensuring cross-session memory notes are correctly stored and interconnected.

Can I use a local markdown wiki to manage knowledge retrieval and linting for software projects?▼

A local markdown wiki supports knowledge retrieval and linting for software projects through built-in automated workflows. It handles conflict resolution, schema management, and pre-computes cross-references so agents can traverse project documentation via link crawling.

When should I not use a markdown wiki for AI agent context storage?▼

A markdown wiki for AI agent context is not ideal for non-text-heavy data or environments lacking local SQLite support. If your project does not require persistent cross-session memory or interlinked documentation, simpler context injection methods may suffice.

Related Skills

Explore More Software Engineering Skills→
Goblin1024Goblin1024
6

wiki

Store and retrieve project knowledge in a Markdown wiki under .omk/wiki with cross-reference links and searchable pages.

Community
Intermediate
ElbertePlinioElbertePlinio

wiki

Generate structured markdown wiki pages with Mermaid diagrams from codebases.

Community
Advanced
zephyrwang6zephyrwang6
333

feishu-wiki

Save Markdown documents to Feishu wiki and read bitable tables via Python scripts.

Community
Intermediate
ricablericable
1

Claude Flow Memory

Manage agent memory with hybrid SQLite and AgentDB backends using HNSW indexing.

Community
Intermediate