brainmaxxing — GitHub Copilot plugin: quizzes and setup
The brainmaxxing plugin quizzes developers on concepts found in their own codebase and builds a competency report. Use it to check what you can explain after AI-assisted coding.
MIT ★ 3 updated 2026-08-19
copilot plugin install brainmaxxing@awesome-copilotWhat brainmaxxing does
brainmaxxing scans a project, identifies the knowledge areas represented in its code, and lets you choose one for assessment. The agent then asks 5 questions tied to actual implementation details. Prompts cover explaining code, predicting behavior, weighing refactoring choices, and reasoning through a bug.
Each answer receives a 0–3 score. After the session, you can compile a report with competency results and study recommendations. The plugin also includes BrainMax Canvas, where you select a domain, submit freeform answers, track scores, and review the report.
The bundled skills cover APIs, databases, architecture, testing, security, DevOps, frontend work, and other domains. They are read-only and never modify the codebase; if Canvas does not open, launch BrainMax once from the Copilot App.
What it helps with
- check whether you understand AI-generated project code
- spot knowledge gaps across implemented domains
- reason about APIs, databases, or tests in context
- score answers against a 0–3 rubric
- compile a competency report with study recommendations
Example prompts for the agent
- Run
/brainmaxand identify the knowledge domains implemented in this project. - Run
/brainmax, choose testing, and quiz me on the repository’s actual tests. - Run
/brainmaxand compile my competency report after the quiz.
How to install brainmaxxing
Command from the source
copilot plugin install brainmaxxing@awesome-copilotStep by step, with commands for each platform: Claude Code plugins: marketplaces, installing, and building your own
Questions about brainmaxxing
How do I install brainmaxxing in GitHub Copilot?
Run copilot plugin install brainmaxxing@awesome-copilot, start Copilot in the project you want to assess, then invoke /brainmax.
What does brainmaxxing do?
It examines the codebase, routes you to a relevant domain quiz, and asks 5 questions grounded in the project. Answers are scored from 0 to 3, followed by a competency report.
What should I do if BrainMax Canvas does not open?
Open BrainMax once from the Copilot App after invoking /brainmax. The current quiz state will then stream into Canvas.
Source text
SKILL.md · MIT
# BrainMaxxing
Agent skills that quiz students on software engineering concepts grounded in their actual codebase. Designed for early-career developers who co-create projects with AI and need to verify their conceptual understanding.
## The problem
AI coding agents make it easy to build sophisticated software without fully understanding the concepts behind it. BrainMaxxing helps students verify whether they actually understand what's in their project — or if they're just running code they can't explain.
## How it works
1. **Invoke `/brainmax`** in your project
2. The agent analyzes your codebase and detects which knowledge domains are implemented
3. You pick a domain (API Design, Database Design, Testing, etc.)
4. The agent quizzes you with 5 questions grounded in *your actual code*
**Rooted in [Bloom's Taxonomy](https://en.wikipedia.org/wiki/Bloom%27s_taxonomy)**, questions escalate through cognitive levels:
- **Explain** (Remember/Understand) → *describe what this **code does** and **why it exists***
- **Predict** (Analyze) → ***trace behavior** through code without running it*
- **Refactor** (Apply/Evaluate) → *identify **tradeoffs** and **improve** what's there*
- **Debug** (Analyze/Evaluate) → *spot a real issue and **reason about its consequences***
5. Each answer is scored on a 0–3 rubric
Evaluation and scoring mirrors the progession above:
- 0 → no recognition
- 1 → names it but can't apply it
- 2 → explains and reasons about change
- 3 → evaluates tradeoffs and could teach it.
6. When done, "compile report" to get your competency summary with personalized study recommendations
<a href="https://gh.io/brainmaxxing/demo">
<img src="assets/brainmaxxing-poster-dark.png" alt="BrainMaxxing Demo Preview" width="800" />
</a>
## Skills
| Skill | What it quizzes |
|-------|----------------|
| `/brainmax` | Orchestrator — detects domains, routes to quizzes, compiles report |
| `/api-design` | REST/GraphQL/gRPC design, HTTP semantics, middleware |
| `/database-design` | Schema design, normalization, queries, migrations |
| `/system-architecture` | Service boundaries, communication, infrastructure |
| `/implementation-patterns` | Design patterns, abstractions, code organization |
| `/testing-strategy` | Test design, mocking, coverage, test pyramid |
| `/security-fundamentals` | Auth, validation, encryption, OWASP concepts |
| `/devops-and-ci-cd` | Pipelines, deployment, infrastructure as code |
| `/error-handling-and-resilience` | Retries, circuit breakers, graceful degradation |
| `/requirements-and-scope` | Requirements engineering, scope management |
| `/domain-modeling` | DDD concepts, aggregates, business rules |
| `/ui-and-frontend` | Component architecture, state, rendering, accessibility |
| `/observability` | Logging, metrics, tracing, health checks |
## Installation
Install the complete BrainMaxxing plugin to get all skills and the interactive BrainMax Canvas:
```bash
copilot plugin install juliamuiruri4/brainmaxxing
```
Then start Copilot in the project you want to assess and invoke `/brainmax`.
### Skills-only fallback
Clients without Copilot plugin or Canvas support can install the same quiz skills without the Canvas:
```bash
npx skills add juliamuiruri4/brainmaxxing
```
Invoke `/brainmax` after installation. The complete quiz flow remains available in chat.
## Maintaining bundled references
Each installed skill is self-contained. Runtime reference files live inside the individual skill directories so `npx skills add` works without needing a repository-level `shared/` directory.
`shared/` remains the canonical authoring source for shared quiz/orchestrator reference material. After editing a file in `shared/`, sync the bundled copies and verify they match:
```bash
./scripts/sync-skill-references.sh
./scripts/sync-skill-references.sh --check
```
## Design principles
- **Read-only**: Skills analyze code, never edit it
- **Grounded**: Every question references specific code in the student's project
- **Honest**: Never fabricates bugs, issues, or code that doesn't exist
- **One-at-a-time**: Questions are asked sequentially, not batched
- **Rubric-scored**: All answers scored on a consistent 0–3 rubric
## Canvas extension
The BrainMax Canvas is included with the plugin and launches when `/brainmax` publishes its detected domains. Students can select a domain, read each question, submit freeform answers, follow their score, and review the compiled report from the canvas. Answers are relayed into the Copilot session; question generation and rubric scoring remain with the domain skills.
If the host does not open the canvas when `set_domains` is invoked, open BrainMax once from the Copilot App. The current quiz state will then stream into it automatically. Skills-only installations continue entirely in chat.
## License
MIT
© Julia Muiruri · MIT · juliamuiruri4/brainmaxxing