AI agents are becoming long-running software.
Trigora is building the execution layer they need—so agents can use tools, wait for humans, coordinate with other agents, survive failures, and continue working for days.
Durable execution, without replay.
https://t.co/m1sYS8zs8c
I built an interactive recovery lab for Transparent Continuation Checkpointing (TCC).
Run a program for up to 25,000 steps, crash it, and watch TCC restore it directly from its latest saved state—without replaying the steps that came before.
https://t.co/fMbVZcbCIn
Long-running AI agents need durable execution.
Our first blog post covers the architecture, early results, and what remains difficult:
https://t.co/QLnmkAzLSX