Most agent bugs are invisible: the agent loops, drops context or calls a tool it shouldn't, and a...Most agent bugs are invisible: the agent loops, drops context or calls a tool it shouldn't, and a...
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
Most agent bugs are invisible: the agent loops, drops context or calls a tool it shouldn't, and all you see is a wrong answer. I wrote an agent harness from scratch in Python — loop control, context compaction, permissioned tools, sub-agents — and Harness Studio, a web app that replays every run: one lane per agent, each event explained, and the exact context each agent held when it stopped.
Post image
Eric's avatar
The context each agent held when it stopped is the part I'd use most. I'd mark the first unexpected tool call in the replay too, so I can see where the run went off course.
Back to feed
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