Run the existing health-check package, copy the safe summary, then use the template below to turn the result into one useful next fix.
Run the check locally
npx @iwo-szapar/second-brain-health-check --no-openThe health check is meant to inspect your local setup. Keep secrets, raw customer data, private repo code, and API keys out of anything you paste elsewhere.
Overall score, setup/usage/fluency scores, brain maturity, CE patterns, and top fixes.
Score panel, weakest dimension, CE radar summary, and guided fixes.
Manifest output if the user ran the health check in manifest mode.
Fallback when the user cannot run the tool: role, current setup, workflow leak, urgency, and what repeats every week.
This works in ChatGPT, Claude, Gemini, or Claude Code. It asks for the state of the setup, not your private files.
# Second Brain Setup Health Check
## My context
Role:
Audience type: founder/operator | consultant/agency | overloaded operator | other
Current AI setup:
Biggest workflow leak:
Urgency:
Best source to attribute this to: LinkedIn | newsletter | resource page | email signature | partner | other
## Health check output
Overall score:
Setup Quality:
Usage Activity:
AI Fluency:
Brain maturity:
Context engineering patterns:
Top recommendations:
Accepted input format: text report | dashboard summary | brain-manifest.yaml | manual notes
## What I need
1. Name my current Second Brain maturity.
2. Point every claim to a health-check signal or intake answer.
3. Tell me the missing context or memory layer.
4. Identify the workflow leak that should become a skill, checklist, hook, or review loop.
5. Give me the next 3 actions in order.
6. Route me: no offer yet, self-serve, Kickstart, DWY, or MemoryOS.
## Privacy confirmation
I removed secrets, customer data, raw private repo code, and API keys before sharing this.Use the getting-started path. Do not route straight to DWY unless pain and urgency are explicit.
Point to MemoryOS or self-serve setup basics. The job is building the spine first.
Use the Context Engineering path and fix the weakest missing layer.
Diagnose the missing CE pattern and recommend the first high-impact fix.
Route to MemoryOS, Team Brain, or advanced support. Starter advice will feel too basic.
Share the smallest useful summary. The goal is diagnosis, not a full repo review.
If the check says your context, memory, or review loop is weak, adding more agents usually makes the setup noisier. Stabilize the spine first, then automate.