Intel

AIKIDO-2026-584771

strands-agents-tools is vulnerable to Improper Access Control

Improper Access Control Pre-CVE
Found by Aikido Intel before public disclosure or CVE publication.
Published 2 days ago

65

Medium Risk

This Affects:

PYTHONstrands-agents-tools
0.2.0 - 0.8.2
Fixed in 0.8.3
Are you affected? Scan for Free

TL;DR

The elasticsearch_memory, mongodb_memory, and mem0_memory tools exposed tenant-identity parameters (namespace, user_id/agent_id) and, for mongodb_memory, the connection string as fields the large language model could set through the tool schema. Because these values determine which tenant's memories are read or written, a crafted prompt could point the tool at another tenant's namespace and read or modify their stored memories, and could redirect the MongoDB connection to disclose its credentials. The values were validated as identifiers but were never bound to a trusted principal. The fix moves tenant identity and connection settings into per-principal tool configuration and environment variables so they are no longer agent-controllable.

Who does this affect?

You are affected if you are using a version that falls within the vulnerable range and use one of the memory tools (elasticsearch_memory, mongodb_memory, or mem0_memory) in a multi-tenant deployment.

Background info

strands-agents-tools is vulnerable to Improper Access Control in versions 0.2.0 - 0.8.2.

How to fix this

Upgrade the strands-agents-tools library to the patch version.