Cohere publishes traces of Qwen3.8 emitting odd URLs in agent commands
Source
huggingface.co
Author
Cohere
Date
Why it matters
If an AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → runs a model that writes unexpected download or upload URLs in shell commands, it can send data to or fetch payloads from untrusted hosts. The released trace lets you test your own agent harnessThe scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.Full definition → and add command allowlists and egress controls.
Terms in this piece · Glossary
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
agent harness — The scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.
