> ## Documentation Index
> Fetch the complete documentation index at: https://opensre.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Closed-Loop Learning

> How the agent carries lessons from one session into the next

# Closed-Loop Learning

OpenSRE learns across sessions through [long-term memory](/docs/platform/memory). When a
conversation surfaces a durable fact — a root cause worth remembering, a
known-flaky service, a naming convention, a preference — the agent stores it
as a local memory file, and the memory index grounds every future
conversation.

There is nothing to configure: memories accrue as you work. Review what the
agent has learned with `/memory`, inspect one entry with `/memory show <name>`, and correct course by deleting stale entries with `/memory forget <name>` or by simply telling the agent the updated fact.

See [Long-Term Memory](/docs/platform/memory) for storage locations, gateway scoping, and
how to turn memory off.