Online · self-paced · 14 chapters
Vibe Coding Security
This course is for people who already build or ship software with AI coding tools. Learn how to review the parts that protect users, money, data, and your production environment, then turn the findings into concrete release checks.
By Teslim Salawu. Read the lessons, check your understanding, and return to the sections you need.
Pilot · paid enrollment not open
Classroom access
Existing Slim Prep students can request instructor-granted pilot access with a verified email address. Registration alone does not enroll you.
Self-paced course: USD 49. Included in the USD 99 bundle with AI Product Management and AI Governance. Paid enrollment remains closed pending launch approval. Optional live teaching is separate.
Ask about pilot accessThe classroom is being prepared at learn.theslimprep.com.
Classroom links will be added after the learning app is deployed. This page is the public course overview.
Course outcomes
What you will learn to do.
- Map what your app protects, where trust changes, and which actions an AI agent may take.
- Keep secrets out of browser code and test sign-in, account recovery, and session behavior.
- Check ownership and tenant boundaries on the server for every protected operation.
- Review data handling, validation, browser protections, uploads, and outbound requests.
- Verify payment events and access grants without trusting a success page or browser claim.
- Prepare deployment controls, security tests, rollback steps, and an incident response plan.
Audience
Who the course is for.
- Founders and independent builders already using AI coding tools to build or ship web applications.
- Product practitioners who can run and inspect a working prototype.
- Developers reviewing the security of an AI-assisted application.
Before you begin
You should be able to run an app locally, read basic JavaScript or TypeScript and HTTP requests, and use environment variables and Git. Examples use Next.js, React, TypeScript, PostgreSQL with node-postgres, the Stripe Node SDK, and GitHub Actions. Each explains the control to look for in another stack. Practice is a design walkthrough unless you bring an application and isolated environment you own; a runnable PocketDesk starter app is not included. Optional executable labs use tools such as Docker Compose, Mailpit, and the Stripe CLI in test mode. Security experience is not required.
How you learn
Work through written lessons and automatically graded chapter assessments. Single-select and multi-select questions check your understanding. Feedback points to relevant lesson sections, so you can review and try again. Practical exercises are self-guided; no live attendance or instructor-graded submissions are required.
The curriculum
14 chapters, in sequence.
The sequence builds from foundations to decisions and operating practice. You can revisit earlier lessons as you work.
Chapter 01
Map what your coding agent can touch
Identify sensitive assets, trust boundaries, agent permissions, and useful limits.
Chapter 02
Protect credentials in source, builds and logs
Separate public configuration from server secrets and prepare for exposure.
Chapter 03
Secure registration, sign-in and recovery
Verify identity, protect sessions, and make account recovery resist abuse.
Chapter 04
Enforce permissions for each workspace
Enforce permissions and tenant boundaries on each sensitive server operation.
Chapter 05
Protect data from collection to recovery
Control collection, access, derived copies, retention, deletion, and logs.
Chapter 06
Validate input and keep data out of commands
Validate shapes and limits, use safe queries, and distinguish data from executable instructions.
Chapter 07
Protect browser content, sessions and private data
Understand rendering, cookies, cross-origin requests, and browser security controls.
Chapter 08
Review dependencies and agent tools
Review packages, tool servers, installation permissions, and changes to the supply chain.
Chapter 09
Restrict file uploads and URL previews
Bound uploads, isolate storage, and prevent unsafe server-side requests.
Chapter 10
Grant paid access from verified payments
Verify payment events, handle retries, and preserve the correct entitlement state.
Chapter 11
Limit AI access, actions and spend
Limit model inputs, tools, output handling, budgets, and abusive use.
Chapter 12
Limit deployment access and respond to alerts
Separate environments and permissions, monitor meaningful signals, and plan safe rollback.
Chapter 13
Test security controls before release
Collect repeatable negative-test evidence and make an explicit release decision.
Chapter 14
Contain incidents and verify recovery
Limit exposure, preserve useful records, repair the cause, and verify recovery.
Optional live class
Learn with an instructor.
The core course remains fully self-paced. Express interest in a separate live Vibe Coding Security class; dates, format, availability, and fees will be confirmed separately. Live teaching is not included in the course or bundle price, and an inquiry does not book a session.
Ask about a live Vibe Coding Security class →About the instructor
Teslim Salawu
Teslim Salawu has spent more than a decade turning problems into products. His experience spans product management, AI, and AI governance. He is the author of AI Product Management.
Before you enroll
Course questions
Is this the same as the live Vibe Coding program?
No. The live program guides a small cohort through building a product. This self-paced course focuses on reviewing the security of an application built with AI tools. The self-paced price includes feedback on your final course project. It does not include live build sessions or an independent security audit of your app. Optional live classes can be discussed separately.
Can I join an optional live class?
You can express interest in a separate live Vibe Coding Security class. Dates, format, availability, and fees will be confirmed separately. The USD 49 course and USD 99 bundle cover self-paced access, not live teaching. The contact link does not book a session or grant course access.
Do I need to use a particular AI coding tool?
No particular tool is required. The course focuses on application behavior, evidence, and trust boundaries that you can inspect across coding assistants. Tool access, subscriptions, and hosting are separate from course access.
How are the lessons assessed?
Each chapter has automatically graded single-select and multi-select questions. Feedback directs you to relevant sections to revisit, and you can retake an assessment. Build a PocketDesk release-review dossier through the chapters and send it for instructor review. Passing every chapter assessment and receiving project approval are both required for completion. You can revise the project if changes are requested.
Will completing the course prove my application is secure?
No. Passing the assessments records your understanding of the course material. It is not a security audit, penetration test, or guarantee about a particular application. The practical work helps you identify evidence, failures, and questions that need further review.
Can I buy the course or use the USD 99 bundle?
The self-paced course is USD 49, or USD 99 for all three courses: AI Product Management, AI Governance, and Vibe Coding Security. Paid enrollment is not open during the pilot. Existing Slim Prep students can request instructor-granted access using a verified email address; account registration alone does not grant access.
