memory

Store and retrieve session context with ChromaDB embeddings.

110|14|Updated May 12, 2026

Search Tags:#semantic search #memory #project scoping #context recall #agent tooling #session summaries #chromaDB

Ready to run

npx skills add https://github.com/EliasOulkadi/shokunin --skill memory-eliasoulkadi

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: memory
Source: https://github.com/EliasOulkadi/shokunin/tree/main/.pack/skills/memory
Command: npx skills add https://github.com/EliasOulkadi/shokunin --skill memory-eliasoulkadi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents you from re-explaining prior work by giving your AI persistent, searchable memory of past conversations, decisions, and code context.

Core Features & Use Cases

Semantic memory across sessions: Stores conversation context in a local ChromaDB vector store and retrieves it at the start of new sessions.
Task-oriented retrieval: Helps you recall “what we decided” for a specific project/topic and avoids mixing unrelated context.
Safe, practical workflows: Encourages storing summaries and checkpoints (not every message), supports forgetting specific entries, and includes guardrails against irrelevant/overbroad queries.

Quick Start

Ask: "Remember that we agreed on the auth refactor approach and that session 7 used verify_file_path to validate file paths."

Frequently Asked Questions about memory

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

FAQPage Schema

How do I enable persistent semantic memory for AI agent conversations across sessions?▼

Persistent semantic memory for AI agent conversations is enabled by storing and retrieving session context with ChromaDB embeddings, allowing you to recall decisions and reuse prior solutions across separate conversations.

What is the best way to recall decisions and prior discussions from previous AI sessions?▼

The best way to recall decisions from prior discussions is using query-time similarity search with project and topic scoping, which retrieves relevant past session summaries from a local vector store without mixing unrelated context.

How do I store and retrieve session context using ChromaDB for AI workflows?▼

You store and retrieve session context in ChromaDB by saving task-oriented summaries and checkpoints with metadata-tagged storage, then retrieving them via similarity search when starting new AI agent workflows.

Do I need a local ChromaDB setup to enable context recall for project scoping?▼

Yes, local ChromaDB persistence is required to enable context recall for project scoping, as it provides the vector store infrastructure needed for metadata-tagged storage and similarity search.

Can I delete or forget specific entries from my AI session summaries to avoid irrelevant context retrieval?▼

Yes, you can delete specific entries from session summaries, as the system supports forgetting individual records and includes guardrails against irrelevant or overbroad queries during context retrieval.

Why does my AI agent fail to remember decisions from earlier sessions without vector storage?▼

AI agents fail to remember decisions without vector storage because they lack persistent semantic recall, making local ChromaDB embeddings and session-level saving necessary to retrieve past interactions.