What's New - February 2026

The February 2026 update is the biggest yet. Here's everything that's new.

Sellable Services Expansion

The brain now includes 11 sellable service templates (up from 5), each with worked examples, pricing guidance, and client-ready deliverables.

New services added:

Landing Page Audit ($200-600) - CRO analysis with prioritized recommendations
Customer Avatar Council ($300-1,000) - AI-powered customer research
Voice of Customer ($400-1,200) - Analyze customer language for marketing
Product Feed Optimization ($500-2,000) - Structured data improvements
AI-Ready Catalog Audit ($800-1,500) - Prepare product data for AI
Agentic Storefront Optimization ($1,500-2,500) - Advanced AI commerce

Also included:

51 worked examples across all services
Updated how-to-sell guides (3 approaches)
Portfolio website template showing all 11 services

Expert Council + Expert Followup

The Expert Council skill now auto-saves all deliberation outputs to data/council-outputs/ with full metadata.

New: /expert-followup - After a council deliberation, continue the conversation with a specific expert. Load their persona and dive deeper into their perspective.

/expert-followup

/upgrade Command

Check for new Claude Code features and get suggestions for brain improvements:

/upgrade

This reads Claude Code changelogs and recommends settings or features you might want to enable.

Activation Profile Template

A new template for deeper self-understanding at context/business/activation-profile.md. Based on the TLOS framework, it captures:

Problems you see in the world
Problems in your day-to-day work
Your capabilities (what you can DO)
Your identity beyond your job title
What you'd automate if you could

The alien test: "Imagine an alien with a clipboard asks 'Who are you? What do you believe? What are you trying to change?' Most people describe their job. What would YOU say?"

Fill this in to give your brain a much richer understanding of who you are.

Permission to Fail

A new instruction in your brain's CLAUDE.md that explicitly tells Claude to:

Say "I don't know" when it doesn't know
Never make up plausible-sounding answers
Suggest alternatives when it can't do something
Ask for help when it needs more context

This reduces AI hallucination and makes your brain more trustworthy.

Persistent Task Lists

Complex multi-step work (audits, content creation, multi-file projects) now survives session restarts. Tasks persist across conversations.

To enable, ask Claude:

Enable persistent task lists for my brain

Updated Research Libraries

Fresh indexes from newsletters and YouTube transcripts covering the latest AI developments, marketing insights, and business strategy.

How to Get the Update

Existing brain users:

/update

New users: Follow the Getting Started guide to set up from scratch.