ingest

Build and maintain a schema-validated knowledge graph from project sources.

Updated Jul 1, 2026
One-click install
npx skills add https://github.com/memstead/memstead --skill ingest-memstead
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ingest
Source: https://github.com/memstead/memstead/tree/main/plugins/claude-code/skills/ingest
Command: npx skills add https://github.com/memstead/memstead --skill ingest-memstead

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill solves the problem of keeping AI agent knowledge synchronized with project sources without database drift or vendor lock-in by enforcing a schema-validated model directly in your git repository.

Core Features & Use Cases

  • Schema-Validated Ingestion: Automatically builds typed entities from codebases, git history, or documentation, ensuring the knowledge graph remains consistent.
  • Resumable Batch Processing: Handles large data sources through focused, resumable batches that are perfect for running on a loop.
  • Use Case: Use this to maintain a living technical specification or decision log for a software project where the agent automatically updates the project graph whenever the underlying source code or documentation changes.

Quick Start

Run the ingest skill to process the next due binding and update your project memory.

Frequently Asked Questions about ingest

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

FAQPage Schema
How do I build a schema-validated knowledge graph from my codebase and documentation?

To build a schema-validated knowledge graph, you can run the ingest skill to process project sources like codebases or documentation. It automatically builds typed entities and enforces data integrity through pinned schemas and git provenance on every mutation.

How does agent memory synchronization work without database drift?

Agent memory synchronization works by routing ingestion tasks through a CLI-driven projection engine that enforces a schema-validated model directly in your git repository. This approach prevents database drift and avoids vendor lock-in.

Can I handle large codebases for knowledge graph ingestion without losing progress?

You can handle large codebases using resumable batch processing. The ingestion mechanism processes focused batches that are perfect for running on a loop, ensuring you maintain progress even with extensive data sources.

What is the best way to maintain a living technical specification that updates with source code?

The best way to maintain a living technical specification is to use an ingestion skill that automatically updates your project knowledge graph whenever the underlying source code or documentation changes, keeping your decision logs accurate.

Do I need an external database to store my AI project's knowledge graph?

You do not need an external database to store your knowledge graph. The system enforces a schema-validated model directly within your git repository, effectively eliminating database drift and vendor lock-in.