MemGuard, explained for beginners
MemGuard screens what an AI agent stores in memory: known attacks are blocked, and anything that might instruct the agent waits for a person.
What it is
Agents with memory carry notes between sessions. An attacker can plant a note today that steers the agent next month.
MemGuard scans each memory before it is stored. It deliberately errs on the side of asking a person, so it holds some ordinary notes too.
An everyday comparison
A mail room for an agent's diary: known scams are shredded, and anything that reads like an order goes to a manager.
How it works, in 3 steps
- The agent wants to save a memory.
- MemGuard scans it.
- Known attack: BLOCK. Might instruct the agent: HOLD_FOR_REVIEW. Otherwise: ALLOW.
Try it now
Open the five live demo tools → Every button sends a real request to the live product and shows the real answer.
See it on the homepage: MemGuard section →
Who it is for
Teams whose agents keep memory across sessions.
FIRME AI CORPORATION · CAIN-42 is a Byzantine trust fabric. Numbers marked "our own check" were run by us, not by an independent party. Nothing on these pages is a certification.