superhero-tech/super-brain/second-brain-setup

Claude Code and agent skill by superhero-tech/super-brain

Trust tier 2
  • community
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

Use when creating or refreshing a product builder's personal profile in 8-System/about.md - supports both assessment-based (Gallup, DISC, MBTI) and interview-based onboarding

Install

npx vskill@latest install superhero-tech/super-brain/second-brain-setup

SKILL.md

Second Brain Setup

Purpose

Runs a conversation with the product builder and writes the result to 8-System/about.md. That file tells the agent who the human is, how they want to be treated and where they need to be pushed. It works as a pair with 8-System/brain.md, which holds the universal operating rules.

When to use

  • Setting up a new vault for the first time
  • Refreshing the profile after a role change, a new project or a shift in priorities
  • When the AI feels "out of calibration"

How it works

Step 0: Pick a path

Ask:

There are two ways for me to get to know you:
A) I have assessment results - paste your CliftonStrengths (Gallup), DISC, MBTI, 16Personalities, Enneagram, Working Genius or any other personality/competency assessment. I will pull out the parts that change how we work together.
B) Let's talk - I will run a short interview (5-7 minutes).
You can do both: start with the assessment, then fill the gaps by talking.
Path A: Assessment-first
  1. Read the pasted results
  2. Extract ONLY operational information - the parts that change how the AI works:
  • Strengths -> how the AI should frame recommendations
  • Weaknesses and blind spots -> where the AI should actively push
  • Communication style -> how the AI should speak
  • Decision style -> how the AI should present options
  1. Do NOT copy the full report into about.md. The extract is 5-8 bullets in the "How to work with me" section
  2. Ask short follow-up questions about what the assessment does not cover:
  • Current role and context (Q1 from references/questions.md - short version, 2-3 sentences is enough)
  • Six-month goal (Q4 from references/questions.md)
  1. Go to Step 3 (proposal)
Path B: Interview
  1. Ask the questions in references/questions.md in order: Q1 (Context), Q2 (Communication style), Q3 (Development areas), Q4 (Goals)
  2. Rules:
  • Q1: all six sub-questions in one block, one answer
  • Q2.1-Q2.5: five multiple-choice questions, one at a time
  • Q3: multi-select, 2-3 from a list of 8
  • Q4: one open question
  1. Go to Step 3 (proposal)
Step 3: Propose the content of about.md

Do not write the file yet. Show the full proposal:

Here is what I want to write to 8-System/about.md:
```
[full file content]
```
Look right? Say "save" or tell me what to fix.

File format - three sections:

markdown
# About me
## Context
[Role, company, product, users, metrics - from Q1 or from the assessment plus follow-up]
## How to work with me
[5-8 bullets: communication style, strengths, blind spots, how to frame recommendations]
[Path A: extract from the assessment. Path B: from Q2 + Q3]
## Current goal
[From Q4, in their own words]
---
## History
-[YYYY-MM-DD]: profile created [path A: based on CliftonStrengths / path B: interview]
Step 4: Save after confirmation

Only after an explicit "save", "OK", "yes" - write to 8-System/about.md. Append an entry to History.

Re-run mode

When about.md already exists:

  1. Read the existing file
  2. Summarise it: "I already know you a bit. Your current profile says..."
  3. Ask: "What changed?"
  4. Update only the delta
  5. Append a history entry: - [YYYY-MM-DD]: updated [list of sections]

STOP - red flags

  • Writing about.md without showing a proposal and getting confirmation
  • Modifying brain.md (never allowed)
  • Copying a full Gallup/DISC report into about.md instead of an operational extract
  • In re-run mode, asking everything again instead of asking for the delta
  • Asking the user to pick their role from a closed list of job titles
  • Implying that some roles are "more of a product builder" than others

References

  • references/questions.md - the full question set with multiple-choice options

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/superhero-tech/super-brain/second-brain-setup/badge)](https://verified-skill.com/skills/superhero-tech/super-brain/second-brain-setup)

Something wrong with this skill or its page? Tell us.