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
I test AI agents for the failures that demos hide.
Over the past few months I've built Sentinel, a benchmark that puts MCP security guards through 6 classes of attacks in a sandbox, and I evaluate frontier coding agents on real bugs every week.
The biggest thing I've learned: an agent that looks great in a demo can still take unsafe actions, make up an answer, or quietly misread an ambiguous instruction once it meets real users.
If you're shipping an AI feature and want someone to break it before your users do, send me a message. I'll point out 3 issues I'd test first, free.
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
Related posts
This one is for teams that hand everyday work to AI agents. In one view you can see what got automated, how much time it saved, how each agent is doing, and you can ask the AI assistant anything.
In the video, I ask "Who might churn this week?" and Lyra replies with three at risk accounts and check in emails ready to send.
My goal was calm and minimal, with lots of white space, while the screen still feels alive.
What would you add or remove? I'd love your feedback.
I ran a session with my team on how to build a team of AI agents, and how to treat them so they actually deliver.
What we've seen running this setup on client work:
→ Agents running 24/7
→ 3B+ tokens in 6 days
→ A 2-week milestone done in 3 days
The 5 things I taught the team:
Treat agents like teammates, not software.
Give them a clear goal, the right context and tools they're comfortable with.
Tell them the goal, never the solution.
Let the agents propose the options. You make the call.
Talk to one agent, not ten.
A central PM agent takes the goal and coordinates the rest.
Solve today's problem.
Don't build the perfect agent on day one. Start small and let the team grow with the work.
Don't be the bottleneck.
Review at milestones, not every task.
For clients, this changes the pace. My job is understanding the problem and making the decisions. The agents handle the execution.
Highlights from the session 🎥
Building something and want to move faster with an AI-first team? Let's talk.
Two color directions for an AI governance explainer. A draft AI use policy becoming the questions nobody can answer yet.
20 voted
77%
6 voted
23%
26 votes
Closed
blue for sure 💯