Cognitive Impedance Mismatch - Dynamic Mode Switching for AI-Augmented Teams
Stop drowning in Slack. Start shipping again.
When your team generates ideas faster than it can process them, standard consensus breaks down. CIM Pattern gives you three operational modes and tells you exactly when to switch between them.
Your team uses Claude, Cursor, Notion, and Slack. You're shipping code and ideas at machine speed, but you're still making decisions at human speed.
The result:
- 40+ unread Slack channels
- Meetings that end with more questions than answers
- Decision paralysis disguised as "async-first culture"
- Burnout from cognitive overload
- 🟢 Green Zone (
$Z_c < 0.7$ ): Async works - 🟡 Yellow Zone (
$Z_c \ 0.7-1.0$ ): Schedule GUSH session - 🔴 Red Zone (
$Z_c \ge 1.0$ ): Activate The Jam
CIM Pattern is the circuit breaker for cognitive overload.
Option A: Web Calculator
# Open in browser
open tools/calculators/web/index.htmlOption B: Command Line
python tools/calculators/cli/zc_cli.py --interactiveOption C: From Metrics
python tools/calculators/cli/zc_cli.py \
--slack 150 \
--notion 20 \
--ai 30 \
--emails 40 \
--team-size 10| Zc Range | Mode | What To Do |
|---|---|---|
| < 0.7 | 🟢 Study Hall | Keep async. You're good. |
| 0.7-1.0 | 🟡 GUSH | Schedule forced convergence within 48h |
| ≥ 1.0 | 🔴 The Jam | Declare BHO forks + activate BLUES rhythm |
- GUSH (60-90 min): Use this template
- BHO Fork: Use this template
- Slack Integration: Setup guide
Pilot results (12 teams, 90 days, 2025):
- Decision latency: -34%
- Meeting hours: -38%
- Reported overwhelm: -42%
- Team satisfaction: +28%
How we measure:
- Zc tracking via automated metrics
- Pre/post surveys (SUS, NASA-TLX)
- Calendar analysis
- Qualitative team feedback
- Quick Start Guide - Get running in 15 minutes
- Installation - Setup & deployment
- FAQ - Common questions
- Project Structure - 📋 Start here! Complete directory layout
- Architecture - System design & rationale
- Contributing - How to contribute
- Session Continuity - Multi-session development guide
- Vision - 🌟 The big picture, manifesto, philosophy
- Mathematical Foundation - Formal proofs
- Cognitive CRDTs - Distributed consensus for humans
- Protocol Core - Original conceptual framework
- Roadmap - What's next (2026-2028)
- Milestone Status - Current progress tracker
- Changelog - Version history
# Clone repository
git clone https://github.com/pyragogy/protocols.git
cd protocols
# Run automated setup
chmod +x scripts/setup.sh
./scripts/setup.sh
# Edit .env with your API keys
nano .env
# Start everything
chmod +x scripts/run.sh
./scripts/run.shThat's it! Dashboard opens at http://localhost:3000
# Install Python dependencies
pip install -r requirements.txt
# Install Frontend dependencies
cd tools/dashboard/frontend
npm install
cd ../../..
# Copy environment template
cp .env.example .env
# Edit .env with your keys
# Start backend
python tools/dashboard/backend/api.py
# Start frontend (in another terminal)
cd tools/dashboard/frontend
npm start# Copy .env template
cp .env.example .env
# Edit .env with your keys
# Start with Docker Compose
docker-compose up- ✅ Zc Calculator - Web + CLI
- ✅ Slack Notifications - Webhook-based
- ✅ Templates - GUSH, BHO, BLUES
- 🔄 Curator AI - Automated Zc monitoring via Claude API
- 🔄 Notion Plugin - Native Zc tracking in databases
- 🔄 Linear Integration - Mode switching in project views
- 🔄 Discord Bot - Real-time mode recommendations
- Agile: Optimized for predictable work cycles
- CIM: Optimized for variable cognitive load in AI-augmented teams
- Async-first: Static cultural default
- CIM: Dynamic mode switching based on measured Zc
- More meetings: Assumes sync is always the answer
- CIM: Sync only when Zc indicates overload
In other words: CIM is the cognitive operating system for teams using AI tools in 2026+.
- Before: Zc = 1.8, 12h/week in "alignment meetings"
- After: Zc = 0.6, 4h/week in GUSH sessions, BHO forks for deep work
- Result: Shipped 2 major features in 6 weeks (previously stalled for 3 months)
- Before: GitHub PRs with 40+ comments, no decisions
- After: GUSH sessions for contentious RFCs, BHO forks for prototypes
- Result: RFC-to-merge time cut from 45 to 12 days
- Before: Email chains with 100+ replies, decision fatigue
- After: Weekly Zc monitoring, automatic GUSH scheduling
- Result: Strategic decisions made in days instead of weeks
We're building this in the open. Contributions welcome:
- 🐛 Bug reports - Open an issue
- 💡 Feature requests - Discussions
- 🔧 Code contributions - See CONTRIBUTING.md
- 📖 Documentation - Corrections, examples, translations
- 🧪 Pilot teams - Try it and share results
Current focus areas:
- Curator AI implementation
- More language bindings (JavaScript, Go, Rust)
- Integration plugins (Notion, Linear, Discord)
- Academic validation studies
License: MIT - Use this however you want
Citation (if you use this in research):
@misc{terzi2026cim,
author = {Terzi, Fabrizio},
title = {CIM Pattern: Cognitive Impedance Mismatch and Dynamic Mode Switching},
year = {2026},
url = {https://github.com/pyragogy/protocols},
note = {Version 3.0}
}Built on:
- Peeragogy - Collaborative learning frameworks
- Active Inference (Karl Friston) - Predictive processing
- Stigmergy (Francis Heylighen) - Indirect coordination
- CRDTs - Conflict-free replicated data types
- 💬 Discord - Join the conversation (coming soon)
- 🐦 Twitter - @pyragogy (coming soon)
- 📧 Email - protocols@pyragogy.org
- 📝 Blog - pyragogy.org/blog (coming soon)
Weekly Office Hours: Every Friday 3-4pm UTC - [Calendar link] (coming soon)
Q: Is this just "Agile for AI teams"?
A: No. Agile assumes predictable sprint cycles. CIM Pattern adapts dynamically to measured cognitive load.
Q: Do I need to track Zc constantly?
A: No. Check weekly, or when the team feels overwhelmed. It's a diagnostic tool, not a micromanagement system.
Q: What if my team rejects GUSH sessions?
A: GUSH is forced convergence—it's supposed to feel uncomfortable. But if resistance is high, you might have cultural pre-requisites to address first.
Q: Can this work for enterprise (100+ people)?
A: Partially. CIM works best for teams of 8-25. Larger orgs need to break into smaller units.
Q: Where's the AI Curator?
A: Coming in v3.1. For now, use the CLI calculator and manual Slack notifications.
- v3.0 (Feb 2026) - Production-ready tools, templates, documentation rewrite
- v2.0 (Jan 2026) - Mathematical formalization, pilot validation
- v1.0 (Dec 2025) - Initial conceptual framework
This month:
- Curator AI alpha release
- Notion plugin beta
- 5 new case studies
This quarter:
- Academic paper submission
- Linear/Asana integrations
- Community governance model
Created by: Fabrizio Terzi (@BergamoHub)
Contributors: See CONTRIBUTORS.md
Special thanks to:
- Teams who piloted v1 and v2
- Claude (Anthropic) for making this README possible
- Everyone who told us v2 was "too abstract"—you were right
Stop drowning in information. Start shipping again.