using-forgetful-memory

What problem does it solve?

Forgetful semantic memory guidance to decide when to query vs create, how to structure content using atomic memory principles, and how to apply importance scoring for prioritization.

Core Features & Use Cases

• Query-before-create workflow: Always check for existing memories before creating duplicates to avoid clutter.
• Atomic memory principles: Each memory captures a single concept with a concise title and content, enabling clean linking and retrieval.
• Content constraints & quality: Guidance on Title (≤200 chars), Content (≤2000 chars), Context (≤500 chars), Keywords (≤10), and Tags (≤10) to improve clustering and search.
• Project discovery: Find and attach memories to the correct project by discovering the current repo, filtering by repo, and using the project_id.
• Announcing operations: When creating significant memories, announce results to improve visibility and traceability.

Quick Start

1. Get current repo with git remote get-url origin and derive the owner/repo identifier. 2) Discover or create a project with list_projects and use the returned project_id. 3) Before creating a memory, run a query to check for duplicates: execute_forgetful_tool("query_memory", {"query": "<topic of potential new memory>", "query_context": "Checking for existing memories before creating", "k": 5})
2. If a similar memory exists, update or link it; if not, create a new memory with a suitable title, content, context, keywords, and tags.