bRRAInSkills stacks · Practitioner
v2.0

AI Coding with bRRAIn

Run an AI coding assistant on governed bRRAIn memory: MCP-first sessions, engineered context, verified gates, decisions that stick.

Level
Practitioner
Learning time
18 hours
Price
$499
Credential
Valid 3 years
Enroll on learn.brrain.io See the syllabus
What's new in v2.0

What changed in this edition.

  • The session method is now MCP-first: start_brain_session, live files created at open, and record_assistant_turn, record_decision and record_learning as you work. The open.md / closure.md git workflow is taught as the variant for teams with a vault mirror.
  • New hands-on setup for Claude Desktop (.mcpb), the bRRAIn VS Code extension, Claude Code and the CLAUDE.md hookup, plus targeted vault retrieval with search_vault and filetype: filtering.
  • Real limits of large memory files: the 256 KB read cap, how oversized indexes silently truncate, and how to keep Master-AI-Context and indexes usable.
  • Skills, hooks and sub-agents are taught as the Claude Code features they are; model choice is taught as durable tiering principles inside bRRAIn's Handler-default, commercial-opt-in policy.
  • Nexus Memory as the human view of what assistants record, with corrections under the Correction & Supersession standard.
  • Labs and the capstone now run as AI role-plays in your browser, scored by AI against published rubrics; the exam is assembled per candidate from a larger bank with four performance items.
Outcomes

What you will be able to do.

  • You will be able to connect Claude Code, Claude Desktop or VS Code to your organization's vault and run MCP-first sessions that any person or agent can resume.
  • You will be able to engineer four-frame prompts whose Memory is retrieved from the vault, compressed and cited.
  • You will be able to drive AI-written changes through tests, types, schema and audit gates, and catch hallucination, drift, scope creep and premature optimization.
  • You will be able to apply zero-trust invariants to privileged code paths and to your assistant's own credentials.
  • You will be able to compose commands, skills, hooks and sub-agents, and choose model tiers within your organization's model policy.
  • You will be able to record and curate decisions and learnings so that canonical memory stays concise, consistent and correctable.

Who it's for

  • Senior engineers adopting an AI coding assistant and governed memory into their daily flow
  • Tech leads bringing AI-assisted coding into team practice
  • Founders shipping product with a small team
  • Open-source maintainers scaling themselves with AI

Not covered here

  • Multi-contributor and team flow (see AI Team Development)
  • Release train design (see AI SDLC & Release Train)
  • System architecture (see AI Augmented Systems Architect)
  • Building extensions on the Platform SDK (see SDK Developer)
  • Framework-specific tutorials (language and framework agnostic)
Syllabus

7 modules, 61 lessons.

About 18 hours of learning. Open a module to see every lesson.

  1. Foundations: governed memory, the zones and the loop 8 lessons · 2 h 15 min

    Why governed memory changes AI-assisted engineering, the eight-zone architecture from an engineer's seat, the seven-station MCP-first loop, a lean Master-AI-Context and the nine-stage build methodology. Lab 1 audits a bloated context file.

    1. Module 1 pretest Diagnostic pretest · 5 min
    2. Why AI memory matters for engineers Reading · 15 min
    3. The eight-zone architecture for engineers Reading · 15 min
    4. The bRRAIn engineering loop — MCP-first Reading · 15 min
    5. The Master-AI-Context anatomy — and how it bloats Worked example · 15 min
    6. The nine-stage build methodology, applied by an engineer Worked example · 15 min
    7. Lab 1: Audit a bloated Master-AI-Context AI role-play lab · 45 min
    8. Retrieval: 7 questions across Module 1 Retrieval check · 10 min
  2. Session method and workspace initialization 10 lessons · 2 h 45 min

    The canonical memory files; wiring Claude Desktop, VS Code and Claude Code; the MCP-first open and close; a worked session with fading; the shell-and-git variant; POPE tagging. Lab 2 survives a forced restart.

    1. Module 2 pretest Diagnostic pretest · 5 min
    2. The canonical memory files Reading · 15 min
    3. Wiring your tools: Claude Desktop, VS Code, Claude Code and CLAUDE.md Worked example · 15 min
    4. Opening a session: start_brain_session and live files Reading · 15 min
    5. Recording and closing a session Reading · 15 min
    6. Worked example with fading: one session, three times Worked example · 15 min
    7. The shell-and-git variant: open.md and closure.md Reading · 15 min
    8. POPE tagging on the files your sessions create Reading · 15 min
    9. Lab 2: Initialize a project's memory and survive a forced restart AI role-play lab · 45 min
    10. Retrieval: 8 questions across Module 2 Retrieval check · 10 min
  3. Context engineering and vault retrieval 9 lessons · 2 h 30 min

    Four-frame prompts, compressing a large project into a cited Memory frame, targeted retrieval with search_vault and filetype:, checkpoints, NextSteps as the active-task pointer and multi-file ambiguity. Lab 3 drives a multi-file refactor without drift.

    1. Module 3 pretest Diagnostic pretest · 5 min
    2. The four-frame prompt — memory · scope · task · acceptance Worked example · 15 min
    3. Compressing a large project into a Memory frame that survives Worked example · 15 min
    4. Targeted retrieval from the vault Reading · 15 min
    5. The session-checkpoint pattern — when to compress and re-seed Reading · 15 min
    6. Using NextSteps as the active-task pointer Reading · 15 min
    7. Multi-file context — removing ambiguity before the assistant guesses Scenario · 15 min
    8. Lab 3: Drive a multi-file refactor without context drift AI role-play lab · 45 min
    9. Retrieval: 8 questions across Module 3 Retrieval check · 10 min
  4. Verifiable gates, failure modes and privileged paths 10 lessons · 3 h 15 min

    The four acceptance gates; hallucination, drift, scope creep and premature optimization; zero-trust invariants on privileged paths including the assistant's own credentials; audit invariants and bRRAIn's hash-chained audit log. Labs 4 and 5.

    1. Module 4 pretest Diagnostic pretest · 5 min
    2. The acceptance-gate framework — tests · types · schema · audit Reading · 15 min
    3. Hallucination patterns — fabricated APIs, invented constraints, mis-cited docs Reading · 15 min
    4. Drift patterns — silent contradiction with earlier decisions Scenario · 15 min
    5. Doing more than asked: scope creep and premature optimization Scenario · 15 min
    6. Zero trust on privileged code paths — the assistant never sees production secrets Reading · 15 min
    7. Audit invariants — making every AI-touched change accountable Reading · 15 min
    8. Lab 4: Catch and correct three induced failures AI role-play lab · 45 min
    9. Lab 5: Add zero-trust gates to a privileged feature AI role-play lab · 45 min
    10. Retrieval: 9 questions across Module 4 Retrieval check · 10 min
  5. Skills, hooks, sub-agents and model choice 8 lessons · 2 h 45 min

    Custom slash commands with provenance, Claude Code skills and hooks combined with the bRRAIn tools, sub-agents for parallel and context-heavy work, and model tiering inside a Handler-default, commercial-opt-in policy. Labs 6 and 7.

    1. Module 5 pretest Diagnostic pretest · 5 min
    2. Slash commands — when they help and when they hurt Reading · 15 min
    3. Skills and hooks — reusable know-how and deterministic guardrails Worked example · 15 min
    4. Sub-agents — parallel and isolated AI work Reading · 15 min
    5. Choosing models and token budgets — within your organization's model policy Reading · 15 min
    6. Lab 6: Design a three-part pipeline that ships a feature AI role-play lab · 45 min
    7. Lab 7: Defend model and budget choices under policy pressure AI role-play lab · 45 min
    8. Retrieval: 8 questions across Module 5 Retrieval check · 10 min
  6. Memory discipline and triage 8 lessons · 2 h 15 min

    Recording at the moment of decision, the Consolidator pattern, AI-versus-human triage, the 'AI does everything' and 'AI does nothing' anti-patterns, and reviewing and correcting memory in Nexus. Lab 8 consolidates a long session.

    1. Module 6 pretest Diagnostic pretest · 5 min
    2. Recording at the moment of decision Reading · 15 min
    3. The Consolidator pattern — AI records liberally, people curate canonical Reading · 15 min
    4. AI-versus-human triage — cost, risk, novelty, reversibility Scenario · 15 min
    5. Anti-patterns — 'AI does everything' and 'AI does nothing' Scenario · 15 min
    6. Reviewing and correcting memory in Nexus Worked example · 15 min
    7. Lab 8: Consolidate a long, messy session into clean canonical memory AI role-play lab · 45 min
    8. Retrieval: 8 questions across Module 6 Retrieval check · 10 min
  7. Capstone and exam readiness 8 lessons · 2 h 15 min

    The capstone brief and evidence pack, pacing, the retro format, the four anchor exemplars and exam strategy, ending with the capstone: ship a privileged feature with an AI assistant and bRRAIn memory.

    1. Module 7 pretest Diagnostic pretest · 5 min
    2. The capstone brief — what is given and what you submit Reading · 15 min
    3. Pacing the capstone — plan, checkpoints and when to stop Worked example · 15 min
    4. The retro format — attributing cause honestly Reading · 15 min
    5. Anchor exemplars — what 92, 78, 71 and 54 look like Scenario · 15 min
    6. Exam readiness — format, timing and the item types that cost points Reading · 15 min
    7. Retrieval: 7 questions across Module 7 Retrieval check · 10 min
    8. Capstone: Ship a privileged feature with an AI assistant and bRRAIn memory AI role-play lab · 45 min
Labs and capstone

Practice against someone who pushes back.

Labs run in your browser as AI role-plays. An AI plays the person on the other side of the scenario — with their own goals and objections — and your work is scored against the published rubric. There is nothing to install.

  • Lab 1 · Foundations: governed memory, the zones and the loop

    AI role-play with a project owner whose Master-AI-Context has grown very large and whose assistant keeps missing conventions

  • Lab 2 · Session method and workspace initialization

    AI role-play initializing a new project's memory from a brief, running an MCP-first open and surviving a forced session restart

  • Lab 3 · Context engineering and vault retrieval

    AI role-play steering a simulated assistant through a seven-call-site refactor that drifts unless context is engineered

  • Lab 4 · Verifiable gates, failure modes and privileged paths

    AI role-play in which a simulated assistant commits an unannounced hallucination, drift and scope creep

  • Lab 5 · Verifiable gates, failure modes and privileged paths

    AI role-play with a Security Controller reviewing an assistant-built two-factor reset feature

  • Lab 6 · Skills, hooks, sub-agents and model choice

    AI role-play designing a command, skill, sub-agent and hook pipeline for recurring export features, with a mid-run failure

  • Lab 7 · Skills, hooks, sub-agents and model choice

    AI role-play with an engineering manager pushing blanket model choices and a hard-coded commercial provider

  • Lab 8 · Memory discipline and triage

    AI role-play curating a long, meandering session's scratch into canonical memory for a tech lead

  • Lab 9 · Capstone and exam readiness

    AI role-play with a product owner, simulated vault and simulated coding assistant for a privileged tenant due-date feature; evidence pack submitted for AI scoring

Capstone

Ship a privileged feature with an AI assistant and bRRAIn memory

Artefact submitted in the capstone lab, AI-scored against the published rubric

Pass mark: 72%

Scored on

  • Plan and triage15%
  • Session method and memory discipline25%
  • Context engineering20%
  • Gates, failure handling and privileged paths25%
  • Retro and attribution15%
Exam and credential

One exam. A credential anyone can verify.

The exam

Items per form
59
Time allowed
120 min
Pass mark
72%
Performance tasks
4
Attempts included
2
Wait between attempts
7 days
  • Online and timed, taken on learn.brrain.io.
  • Your form is assembled for you from the course's item bank, so no two candidates sit the same paper.
  • Performance tasks are conducted by an AI examiner: you work through a realistic scenario and are scored against a published rubric.

The credential

  • A verifiable digital badge in your name.
  • A public verification page at learn.brrain.io/verify, so an employer or client can confirm it.
  • Valid for 3 years.
  • Renewal: At 3 years by passing the then-current exam; quarterly CE modules keep you current in between
Before and after

Where this course sits.

Stacks well with

Questions

Frequently asked.

Do I need to install anything for the labs?

No. Labs and the capstone run in your browser on learn.brrain.io as AI role-plays: an AI plays the person on the other side of the scenario, and your work is scored against the rubric published with the course.

How is the exam delivered?

Online and timed: 59 items in 120 minutes, on a form assembled for you from the course's item bank. 4 of the items are performance tasks conducted by an AI examiner: you do the work rather than pick an answer. The pass mark is 72%.

What if I don't pass first time?

You have 2 attempts, with a 7-day wait after an unsuccessful attempt. Further exam attempts can be bought for $299 each.

How long is the credential valid?

3 years. You receive a verifiable digital badge with a public verification page at learn.brrain.io/verify, so anyone can confirm it is genuine.

I hold the v1 credential. Is it still valid?

Yes. Credentials earned on v1 remain valid and verifiable at learn.brrain.io/verify. When you renew, you sit the then-current version of the exam.

Can my company enroll a team?

Yes. Firms can buy a certification bundle for $2,999 per firm per year — see the pricing page — or contact us to arrange enrollment for a larger group.

Also in bRRAInSkills stacks

Related courses.

See every bRRAInSkills stacks course · Full catalog

Enroll

AI Coding with bRRAIn

Run an AI coding assistant on governed bRRAIn memory: MCP-first sessions, engineered context, verified gates, decisions that stick.

Enroll on learn.brrain.io Back to all courses