What is the memory-search agent?

A specialized sub-agent (typically running on Claude Haiku) that finds relevant past context and summarizes it for the main agent, keeping costs low and insights focused.

Why is the Total Recall protocol mandatory?

It ensures that every technical decision and insight is recorded as infrastructure, preventing the AI from repeating mistakes or losing context in subsequent sessions.

How does synthesis-first save context tokens?

Instead of feeding the entire chat history into the prompt, it uses compressed 'nodes' that Claude only expands when specific details are required.

When should I use synthesis_create?

You must use it whenever a decision is made, a new pattern is learned, or a significant bug is identified to ensure the knowledge is stored permanently.

Total Recall Memory Protocol

Name: Total Recall Memory Protocol
Author: Cygnusfear

byCygnusfear

Productivity & Workflow

Enforces a synthesis-first memory system to capture and retrieve critical decisions and learnings across development sessions.

About

Total Recall is a high-performance memory protocol designed to solve the problem of context loss in long-term AI development projects. By shifting from raw chat logs to a structured synthesis approach, it captures the 'why' behind every decision and discovery, storing understanding rather than just data. This skill utilizes progressive disclosure to save up to 100x on context tokens, ensuring Claude maintains a persistent, searchable history of your project's architectural choices, bug patterns, and technical insights without overwhelming the model's window.

Key Features

Progressive disclosure workflow (one-liner to full rationale)
Dedicated memory-search sub-agent for cost-effective retrieval
0 GitHub stars
Synthesis-first memory capture for decisions, learnings, and gotchas
Automated context injection via SessionStart hooks
Mandatory recording protocols to prevent project amnesia

Use Cases

Documenting complex bug fixes and edge cases to prevent regression in future sessions
Managing long-running software projects where architectural context must persist across weeks
Optimizing token usage in high-context repositories through compressed synthesis

About

Key Features

Progressive disclosure workflow (one-liner to full rationale)
Dedicated memory-search sub-agent for cost-effective retrieval
0 GitHub stars
Synthesis-first memory capture for decisions, learnings, and gotchas
Automated context injection via SessionStart hooks
Mandatory recording protocols to prevent project amnesia

Use Cases

Documenting complex bug fixes and edge cases to prevent regression in future sessions
Managing long-running software projects where architectural context must persist across weeks
Optimizing token usage in high-context repositories through compressed synthesis