bRRAIn Certified SDK Developer

What the Platform SDK is and where extensions run, getting access through a developer seat, wiring the Go module, constructing the client from the supervisor's environment, and handling secrets and service identity.

1. Module 1 pretest Diagnostic pretest · 5 min
2. What the Platform SDK is, and where your code runs Reading · 15 min
3. Getting access and wiring the module Worked example · 15 min
4. Constructing the client and proving the platform is reachable Worked example · 15 min
5. Secrets hygiene and service identity Reading · 15 min
6. Scenario: choosing the right integration surface Scenario · 15 min
7. Lab 1: bootstrap code review AI role-play lab · 45 min
8. Module 1 retrieval check Retrieval check · 10 min

Reading text, JSON and binary files correctly, preparing images and documents, writing and appending safely with attribution, and managing files with an accurate model of delete.

1. Module 2 pretest Diagnostic pretest · 5 min
2. Reading vault files correctly Worked example · 15 min
3. Working with images and documents Worked example · 15 min
4. Writing and appending Worked example · 15 min
5. List, stat, move, copy and delete Reading · 15 min
6. Scenario: the vanishing drafts folder Scenario · 15 min
7. Lab 2: vault operations review AI role-play lab · 45 min
8. Module 2 retrieval check Retrieval check · 10 min

Finding knowledge with Search, Grep and the search_vault tool as they behave on current pods, committing incoming content through policy-aware ingestion, tagging with POPE v1.1, and designing vault paths that respect namespaces and per-project permissions.

1. Module 3 pretest Diagnostic pretest · 5 min
2. Finding knowledge: Search, Grep and the search tool Worked example · 15 min
3. Ingestion commit: governed writes for incoming content Worked example · 15 min
4. POPE tagging at write time Reading · 15 min
5. Designing where your extension writes Reading · 15 min
6. Scenario: ingest or write? Scenario · 15 min
7. Lab 3: ingestion design with a product owner AI role-play lab · 45 min
8. Module 3 retrieval check Retrieval check · 10 min

Model selection with the Handler as default and commercial models as an organization's opt-in, single- and multi-turn calls, robust reply handling, grounded answers with checked citations, and image and document attachments.

1. Module 4 pretest Diagnostic pretest · 5 min
2. The LLM Registry and the Handler Reading · 15 min
3. Calling models: single-turn, multi-turn and replies Worked example · 15 min
4. Grounding answers in vault content Worked example · 15 min
5. Sending images and documents to models Worked example · 15 min
6. Scenario: "just use a commercial model" Scenario · 15 min
7. Lab 4: designing a grounded answer feature AI role-play lab · 45 min
8. Module 4 retrieval check Retrieval check · 10 min

Calling the pod's MCP tools, reading the integration catalog and payload schemas, handling webhook subscription as it works today, and obtaining vendor credentials at call time through OAuth and Operator ENV.

1. Module 5 pretest Diagnostic pretest · 5 min
2. Calling MCP tools from an extension Worked example · 15 min
3. The integration catalog, payload schemas and webhooks Worked example · 15 min
4. Vendor credentials: OAuth bearers and Operator ENV Reading · 15 min
5. Scenario: credential handling review Scenario · 15 min
6. Lab 5: designing a connector-backed feature AI role-play lab · 45 min
7. Module 5 retrieval check Retrieval check · 10 min

Policy checks with trusted identity, roles and the Control Plane as they behave today, audit events and notifications, and what namespace isolation and least privilege do and do not provide.

1. Module 6 pretest Diagnostic pretest · 5 min
2. Policy checks before you act Worked example · 15 min
3. Roles, permissions and the Control Plane Reading · 15 min
4. Audit events and notifications Worked example · 15 min
5. Namespace isolation and least privilege Reading · 15 min
6. Scenario: the role lookup that was never there Scenario · 15 min
7. Lab 6: governance review with the Security Controller AI role-play lab · 45 min
8. Module 6 retrieval check Retrieval check · 10 min

Classifying SDK errors, deadlines and safe retries with backoff, testing against httptest fakes that reproduce current-pod behavior, and debugging extensions on a pod.

1. Module 7 pretest Diagnostic pretest · 5 min
2. The error model: APIError, transport errors and validation errors Worked example · 15 min
3. Deadlines, retries and backoff Worked example · 15 min
4. Testing against a fake platform with httptest Worked example · 15 min
5. Debugging an extension on a pod Reading · 15 min
6. Scenario: duplicate lines after a retry storm Scenario · 15 min
7. Lab 7: incident triage with an operator AI role-play lab · 45 min
8. Module 7 retrieval check Retrieval check · 10 min

The manifest contract, packaging, install and update under the supervisor, marketplace submission and maintenance, recording decisions in AI-assisted sessions, and the capstone.

1. Module 8 pretest Diagnostic pretest · 5 min
2. The extension manifest Worked example · 15 min
3. Packaging, install and update under the supervisor Reading · 15 min
4. Marketplace submission and maintenance Reading · 15 min
5. Recording decisions in AI-assisted development sessions Worked example · 15 min
6. Scenario: release day Scenario · 15 min
7. Module 8 retrieval check Retrieval check · 10 min
8. Capstone: ship a governed digest extension AI role-play lab · 45 min