brain-capture

Capture and deduplicate high-value Claude session moments into a structured knowledge brain.

Updated Apr 25, 2026
One-click install
npx skills add https://github.com/agnivadc/brain_playground --skill brain-capture
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: brain-capture
Source: https://github.com/agnivadc/brain_playground/tree/main/.claude/skills/brain-capture
Command: npx skills add https://github.com/agnivadc/brain_playground --skill brain-capture

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Manages in-session moments and structured notes by tiering, drafting, and sweeping observations into a centralized knowledge brain, reducing memory leakage and ensuring important decisions are preserved with dedup.

Core Features & Use Cases

  • Tier and classify moments as IMMEDIATE or BATCH to control persistence.
  • Draft artifacts in .brain-drafts/ for later review and consolidation.
  • Sweep and integrate final conclusions into the brain when appropriate.

Quick Start

Start brain capture for this session and tier, draft, and sweep notes into the knowledge brain.

Frequently Asked Questions about brain-capture

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

FAQPage Schema
How do I capture high-value decisions from Claude sessions into structured notes?

You can capture high-value decisions from Claude sessions by tiering moments as IMMEDIATE or BATCH, routing them into a structured knowledge brain with deduplication to prevent memory leakage and preserve critical rules.

What is the best way to manage in-session knowledge drafts for later review?

Managing in-session knowledge drafts is handled by routing BATCH observations into a .brain-drafts/ directory, holding them for later consolidation before sweeping final conclusions into the centralized knowledge brain.

How does deduplication work when logging inline rules into a knowledge base?

Deduplication when logging inline rules works by routing captured moments through a structured knowledge brain that applies strict frontmatter-based metadata and sweep workflows, ensuring identical observations are consolidated.

Can I use this knowledge capture workflow for trading-session interactions without external dependencies?

Yes, you can use this knowledge capture workflow for trading-session interactions without external dependencies, as it operates entirely within the Agent OS runtime and requires no additional packages.

When should I tier a captured moment as IMMEDIATE versus BATCH?

You should tier a captured moment as IMMEDIATE for critical decisions requiring instant persistence, and as BATCH for general observations to draft into .brain-drafts/ for later review and sweeping.