A vibecoded SaaS leaked user data because nobody owned the design
A production SaaS built by vibe coding failed in four different ways with one root cause. What survived shows why AI-generated code needs someone who owns the system design.
Written by
-
Benedikt Stemmildt Founder & Co-CEO
Published: 3 March 2026 Updated: 1 min read
Someone vibecoded a production SaaS. The security holes were real: no demo bugs or edge cases, but actual user data leaking between accounts.
What broke
- Row-level security was bypassed, and data leaked across user accounts.
- OAuth flows were corrupted, and authentication became impossible to debug.
- Background jobs shared context between users.
- Environment variables drifted silently between environments.
These are four different categories of failure with one root cause: nobody owned the system design.
What survived
- UI-first development. Let the AI generate interfaces, and own the backend yourself.
- Socratic questioning. Don’t accept the first AI answer. Ask “why?” until the architecture makes sense.
- Documentation as disaster recovery. When everything breaks, good docs are the lifeline.
- An explicit “engineering layer” that owns architectural coherence.
Coherence needs an owner
AI can generate code. It cannot keep architectural coherence across a growing system. That requires understanding the constraints, conventions and boundaries before a single line gets written.
We call this Context Engineering: giving AI structured knowledge about your system so it produces code that fits, and not only code that compiles.
The gap between an AI demo and production reality is discipline, not intelligence.
Are you vibecoding into production, or engineering your context first?
Further reading
Written by
-
Benedikt Stemmildt · Founder & Co-CEO
More on Agent security and risk
- Vouch for the harness, not for its output Code review cannot keep up with agent output. Open source maintainers already admit contributors instead of reviewing contributions, and in an agentic factory the contributor to admit is the harness.
- Well-built agents close compliance gaps that manual processes leave Compliance teams block agents as too risky while manual processes run on memory and reconstructed logs. Properly architected agents make audit trails and enforcement automatic.
- A hook is not a boundary: why agents need a sandbox A training group got an agent past a hook that guarded a secrets file. A hook filters the calls it was written for, a sandbox limits what the agent can reach, and the trade between the two is a company decision.
Test the constraint in your system
The Factory Readiness Assessment connects this research to your value stream, your data, and the next decision.
Training on this topic: Agentic Engineering Security for Development TeamsWant to check one team first? The self-check: twelve yes or no questions, three minutes, instant result, no account. Take the self-check