OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Kipi

Kipi is a Python-based, MIT-licensed framework that externalizes AI memory in plain Markdown and retrieves relevant knowledge for Claude Code, supplying source file and line references with coverage information.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Kipi — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Kipi when a Claude Code workflow needs persistent, question-driven access to project and work knowledge without relying on a vector database, especially when source locations and retrieval coverage matter.

Who it is for

  • Claude Code users who need work context across AI sessions
  • Developers building self-hosted Markdown knowledge workflows
  • Project, research, legal-assistant or investigator workspaces that need source-backed context retrieval

Documented uses

  • Store work knowledge in inspectable Markdown files
  • Select sources relevant to a question
  • Supply retrieved evidence to Claude Code
  • Show the file and line behind retrieved information
  • Report empty, unreadable or partially searched sources
Before you choose

Requirements & limits

The repository is public, active in the supplied record, and licensed under MIT. It is identified as a CLI, desktop and self-hosted framework for the Claude Code ecosystem. No pricing or account requirement is documented.

  • Claude Code is the documented model environment.
  • Knowledge is stored in plain Markdown and can be inspected with standard file tools and Git diffs.
  • Source selection is configurable through q-system/.q-system/knowledge-sources.json.
  • Named sub-stores can scope retrieval to a case or engagement.
  • Coverage reporting distinguishes fully read, empty, unreadable and partially searched sources.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up a role-specific Markdown knowledge workspace

Organize a Kipi knowledge base for a specialized assistant role.

Before you start

  • Claude Code
  • A Kipi checkout from https://github.com/assafkip/kipi-system
  • Plain Markdown knowledge files
  • A project folder using Kipi's documented knowledge-source structure

Steps

  1. Create or organize Markdown knowledge covering the work domain, such as decisions, projects, people and commitments.

  2. Place project material in the documented knowledge folders, including output, research, inputs, investigation, build, docs, notes or memory as relevant.

  3. Use q-system/.q-system/knowledge-sources.json to define which folder names count as document sources.

  4. Use named sub-stores when separate cases or engagements need distinct knowledge scopes.

Expected outcome

The workspace has configurable Markdown sources that Kipi can select according to a question and report with file and line references.

Guide limitations

  • The supplied material documents the knowledge structure and retrieval behavior but does not provide a complete installation or execution walkthrough.
  • A proposed domain schema is not demonstrated as an implemented deployment.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 October 2026
Original website title
assafkip/kipi-system · Official homepage title · retained in its source language
How we review sourcesSuggest a correction