Hermes 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 → v0.21.0, released August 31, 2026, gives scheduled AI jobs a way to carry information between runs. Nous Research's release notes say scheduled agents now load and update persistent memory like other agents, each job gets a durable scratchpad, and a continuity setting passes one run's output into the next.
The notes frame this around monitoring: a job can avoid repeating what it already reported. They also say monitor-mode jobs skip the LLMA large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.Full definition → entirely when nothing has changed. These are documented mechanisms; the release does not report how much repetition or cost they save.
Terms in this piece · Glossary
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.
LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
