The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
A hardened MCP server exposing read-only order tools to an AI agent - treating every tool argument as untrusted, because it is LLM-generated. A naive string-concat server leaks the whole table on a classic OR-1=1 injection; this one blocks 5/5 injection attempts with zero functionality lost. Validation + parameterized SQL + a read-only connection, proven by a contract suite that also demonstrates the naive version really leaks. Tool args are a network boundary - most MCP demos skip that. Public code: github.com/jigonyoo/mcp-server-prod
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Trending
Claude
Claude has entered the design space. How are you using it?
Contra University
Learn from expert creatives how to earn more using next-gen AI tools.
Brand Design
The best brand designers are on Contra. Scroll to see what's trending in brand design. What are you building?
creativeaiflow
Creative AI workflows are evolving. What tools do you use, and what are their strengths and weaknesses?
freelancerlife
Freelancer life is wins, pivots, and everything in between. What’s yours right now?
Related posts
Built Oracle AI, a custom AI architecture and engineering partner, not just another agency selling generic wrappers.
We design and build tailored AI systems to solve real, complex business bottlenecks. Turning fragmented operational problems into production-ready architecture:
Custom AI System Architecture — end-to-end design of scalable AI pipelines, data flows, and infrastructure built specifically for your exact business problem
Autonomous Agent Orchestration — designing agentic systems where specialized models collaborate, execute multi-step workflows, and solve actual tasks
Enterprise Integration & Data Pipelines — connecting cutting-edge AI models directly into your existing codebase, databases, and operational tools
Problem-First Solution Engineering — we don't start with "what model should we use?"; we start with the core problem and engineer the exact system needed to solve it
Production-Ready Security & Reliability — building robust, secure, and compliant architectures ready for real-world traffic, not just flashy prototypes
The point isn't "we can generate cool text with AI." It's that we architect, build, and deploy the underlying intelligence your company needs to solve actual complex operational challenges.
Check it out: oracleai.se
LovableaideveloperAlgorithmic Identity ArchitectureAI Agent DesignerLow-Code/No-Code DevelopmentGoogle GeminiOpenAI
After 157 commits spanning 8 months, one thing became clear...
Just how non-negotiable security is.
I'm not taking about rate limits, or protecting your secrets like API keys, or broken auth, or prevention against XSS or SQL injection, or not committing dotENV to github, or updating privacy policy if you're using any user data, or server side validation... Or anything that comes under OWASP basics.
They have to be there in the first place. From day one.
I refer to subtle race conditions that even an intelligent LLM can sometimes fail to identify. A database leak. Like what happens when a database fails? How does the application handle it? What happens to payments during that period? How to manage similar threats?
Adversaries wait for just one tiny opportunity...
P.S., I'm open to freelance contract or role opportunities. I can build fully secure backend and the frontend system and ship it. If you know of any opportunities, please say hi. Or connect me on LinkedIn : https://linkedin.com/in/wajidfarooq
