Intel

AIKIDO-2026-382117

strands-agents-tools is vulnerable to Server-Side Request Forgery (SSRF)

Server-Side Request Forgery (SSRF)CVE-2026-15746 Published 2 days ago

65

Medium Risk

This Affects:

PYTHONstrands-agents-tools
0.2.11 - 0.6.0
Fixed in 0.7.0
Are you affected? Scan for Free

TL;DR

The elasticsearch_memory tool exposes its Elasticsearch connection parameters (es_url, cloud_id, api_key) as fields the language model can control through the tool schema. When a caller omits api_key, the tool falls back to the operator's ELASTICSEARCH_API_KEY environment variable, and because the destination host is also model-controllable, a crafted prompt can direct the tool to connect to an actor-controlled server. In that case the operator's Elasticsearch credential is sent in the Authorization header to the attacker host, disclosing it. The fix restricts the connection parameters so the model can no longer redirect requests to leak the stored credential.

Who does this affect?

You are affected if you are using a version that falls within the vulnerable range and you use the elasticsearch_memory tool.

Background info

strands-agents-tools is vulnerable to Server-Side Request Forgery (SSRF) in versions 0.2.11 - 0.6.0.

How to fix this

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