Skip to main content

AWS API MCP Server EUVDEUVD-2026-111349

| CVE-2026-115413 HIGH
Origin Validation Error (CWE-346)
2026-10-27 AMZN
7.3
CVSS 4.0 · Vendor: AMZN
View

Attack ChainAIDerived

Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain

  1. Access
    Obtain token from sibling app in same IdP tenant
  2. Delivery
    Reach local streamable-http OAuth endpoint
  3. Exploit
    Present bearer token, audience check bypassed
  4. Execution
    Issue AWS API/CLI calls via host credentials
  5. Impact
    Unauthorized AWS operations within credential permissions

Vulnerability Assessment

Weakness
CWE
OWASP Top 10:2025
A07:2025 - Authentication Failures
ASVS 5.0
V6 - Authentication Origin validation
Exploitation Conditions

Exploitable ONLY when the server is deployed with the non-default streamable-http transport AND OAuth authentication enabled against an operator-nominated identity provider; the default stdio transport is not affected. The attacker must be a local caller (AV:L) who already possesses a valid, unexpired, correctly-signed bearer token issued by that same identity provider to a different application/client in the same tenant (for example a second app client in the same Amazon Cognito user pool) - the server fails to validate the token's audience/client-id claim. Resulting access is bounded by the permissions of the AWS credentials configured on the host. Limiting factors: non-default configuration required, local access required (PR:L / authenticated), and the attacker must obtain a same-tenant sibling-app token (AT:P). Fixed in 1.6.0, which additionally requires operators to set an expected token audience or client ID; the documented workaround is to use the default stdio transport instead of streamable-http.

Impact

This is a genuine but conditional, non-default-exposure issue rather than a drop-everything emergency. The AWS CVSS 4.0 vector (AV:L/AC:L/AT:P/PR:L/UI:N) scores it 7.3 High, and the impact triple is severe (VC:H/VI:H/VA:H) because a successful attacker executes arbitrary AWS operations under the host's configured credentials. However, several signals temper real-world risk: the vector is LOCAL (AV:L), it requires a present attack condition (AT:P) - the attacker must already hold a valid bearer token the operator's own identity provider issued to a different application in the same tenant - and it requires some privilege (PR:L), so it is an authenticated weakness, not anonymous remote code access. No EPSS score and no SSVC exploitation/automatable values were supplied, so probabilistic exploitation likelihood cannot be quantified from this data. The vulnerability is NOT on CISA KEV and no working exploit code is reported; only a vendor/GHSA advisory exists. The vuln.today decision label is 'this_week', consistent with a serious issue that is bounded by non-default configuration (only streamable-http + OAuth deployments are affected) rather than ubiquitous unauthenticated exposure. Net: prioritize patching for the specific affected deployment mode, but this is not a wormable internet-facing critical.

Exploit ScenarioDerived

An operator runs the AWS API MCP Server with the streamable-http transport and OAuth authentication backed by, for example, an Amazon Cognito user pool. A local caller who holds a bearer token that the same Cognito pool issued to a second, unrelated app client in the same tenant presents that token; because the server validates signature, issuer and expiry but not the token's audience/client-id claim, the token is accepted and the caller issues AWS CLI operations using the host's configured AWS credentials. No public working exploit code is reported, though the issue is documented in the awslabs GHSA advisory.

This Week Schedule remediation this week. Why: HIGH severity (CVSS 7.0-8.9).
Patch available 0.1% exploitation probability (EPSS)

Get an e-mail when this CVE is exploited, patched or re-scored. Free account.

Alert me

AnalysisAIv1 – Oct 28, 2026

Missing OAuth token audience validation in AWS API MCP Server 1.3.0 through 1.5.x lets a local caller reach the server using a bearer token that the identity provider issued to a different application in the same tenant, and then run AWS operations with the AWS credentials configured on the host. The flaw is conditional: it only applies to deployments using the non-default streamable-http transport with OAuth authentication enabled against an operator-nominated identity provider, while the default stdio transport is unaffected. Resulting access is bounded by the permissions of the host's AWS credentials; no public exploit code has been identified at time of analysis, and version 1.6.0 fixes the issue and additionally requires operators to configure an expected token audience or client ID.

RemediationAI

Vendor-released patch: upgrade awslabs.aws-api-mcp-server to version 1.6.0 (https://pypi.org/project/awslabs.aws-api-mcp-server/1.6.0/), then configure the expected token audience or client ID for the server - 1.6.0 makes this setting required for OAuth deployments, so an upgrade without it does not complete the fix. Operators who cannot upgrade immediately can switch to the default stdio transport instead of streamable-http, which removes the OAuth token-acceptance path but changes how clients connect to the server and drops the HTTP transport's remote-access capability. Also ensure forked or derivative code incorporates the same change, and review the GHSA advisory (https://github.com/awslabs/mcp/security/advisories/GHSA-4prq-p2x5-w9fh) and AWS bulletin (https://aws.amazon.com/security/security-bulletins/2026-152-aws/) for details; scope the AWS credentials configured on the host to the minimum permissions needed so that any accepted cross-application token yields limited access.

Affected VersionsENISA

aws-api-mcp-server 1.3.0 <1.6.0

Affected Packages

euvd aws_api_mcp_server fix: 1.6.0

Source: OSV (Open Source Vulnerabilities)

Affects your stack

Affects 1 package in euvd

Checking this CVE against your own stacks and servers is part of the Basic plan and up.

Technical ContextAI · show details

The AWS API MCP Server (awslabs.aws-api-mcp-server on PyPI) is an open source Model Context Protocol server that lets AI assistants drive AWS services by executing AWS CLI commands on the host. In its streamable-http transport mode, operators can configure OAuth authentication so that a bearer token presented to the server is validated against an operator-nominated identity provider. Affected versions validate the token's signature, issuer and expiry but never check the claim identifying which application the token was issued for, an origin/audience validation failure classified as CWE-346. Consequently a token minted by the same identity provider for a sibling application in the same tenant - for example a second app client in the same Amazon Cognito user pool - is accepted as valid, and the caller's subsequent AWS API calls execute under the credentials configured on the server host, capped by those credentials' permissions. The exploit path is not a signature or crypto break but a trust-boundary mistake in the resource-server check, so the practical lever is the audience/client-id claim that the software simply ignored; 1.6.0 adds that check and makes configuring the expected audience or client ID mandatory for OAuth deployments.

Know when AWS gets its next vulnerability

  • Watch products and get an email when something new is published
  • Check this CVE against your own stack and servers
  • Exploit and PoC references, threat intelligence from 22 sources

No credit card · 7-day full trial

Severity by source

Vendor (AMZN) PRIMARY 7.3 HIGH
CVSS:4.0/AV:L/AC:L/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
vuln.today AI 7.0 HIGH

Local access and authenticated caller give AV:L/PR:L; needing a same-tenant sibling-app token is an attacker-uncontrolled precondition, hence AC:H; host-credential AWS operations yield C/I/A:H.

3.1 AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H
4.0 AV:L/AC:L/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N

Primary rating from Vendor (AMZN).

CVSS VectorVendor: AMZN

Attack Vector
Local
Attack Complexity
Low
Attack Requirements
Present
Privileges Required
Low
User Interaction
None
Vulnerable system: Confidentiality
High
Vulnerable system: Integrity
High
Vulnerable system: Availability
High
Subsequent system: Confidentiality
None
Subsequent system: Integrity
None
Subsequent system: Availability
None

Lifecycle Timeline

2
CVSS changed
Oct 27, 2026 - 19:22 NVD
7.8 (HIGH) 7.3 (HIGH)
CVE Published
Oct 27, 2026 - 18:10 cve.org
HIGH 7.3

DescriptionCVE.org

Missing token audience validation in the OAuth authentication mode of AWS API MCP Server before version 1.6.0 might allow local users to run AWS operations with the server's configured AWS credentials via an access token issued to another application by the same identity provider.

To remediate this issue, users should upgrade to version 1.6.0 and configure the expected token audience or client ID for the server.

Technical POC Analysis

Evidence caveat: the supplied material is the GitHub advisory GHSA-4prq-p2x5-w9fh plus a PyPI page that failed to render (JavaScript challenge).

Full technical POC analysis - root cause, step-by-step exploitation and patch breakdown - is part of the Pro and Business plans.

Executive ViewAI

Why It Matters

The AWS API MCP Server versions 1.3.0 through 1.5.x contain a HIGH severity flaw (CVSS 7.3) that can allow an attacker with a valid token from a different application to execute AWS operations using the server's host credentials. Only deployments using the non-default streamable-http transport with OAuth enabled are at risk, so leadership should focus on identifying and remediating those configurations. Successful exploitation could lead to unauthorized AWS actions, potentially impacting cloud resources and data.

Business Impact: High
Compensating Controls

Apply vendor patch (version 1.6.0) and configure the expected token audience or client ID.

Threat Intelligence Overview

EPSS Score
0.1%
Top 99% exploitable
Prevalence
Low
niche, recent open-source AWS Labs MCP server; narrow affected deployment mode
Social Media
1
Sightings
Data Quality
4 sources
Confidence: Medium

How this decision was reached

The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.

SignalValue
Severity HIGH
CISA KEV No
Public POC No
EPSS 0.1%
Patch available Yes
Active threat No
Deciding rule
HIGH severity (CVSS 7.0-8.9)
What every action level means
Emergency Immediate action required. Criteria: CRITICAL + in CISA KEV (actively exploited) without patch, OR CRITICAL + EPSS > 30% without patch, OR CRITICAL + active threat signals without patch.
Act Now Prioritize within 24-48h. Criteria: in KEV with patch available, OR CRITICAL + public POC, OR HIGH + EPSS > 10%, OR CRITICAL + active threat, OR any CRITICAL severity.
This Week Schedule remediation this week. Criteria: HIGH severity (CVSS 7.0-8.9). Patch within normal change window.
This Month Plan within 30 days. Criteria: MEDIUM severity (CVSS 4.0-6.9), lower exploitation probability. Include in regular patch cycle.
Monitor Track and reassess. Criteria: LOW severity (CVSS 0.1-3.9), minimal exploitation risk. No immediate action needed.

SSVC AssessmentENISA

SSVC Decision· derived
Defer
Exploitation· derived
None
System Exposure· derived
Small
Automatable· derived
No
Human Impact
High

Priority Score

36
Low Medium High Critical
KEV: 0
EPSS: 0
CVSS: +36
PoC: 0
Weaponized: 0
Threat intel: 0
Gov advisory: 0

Attack Technique

Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.

More in AWS

View all
CVE-2026-103956 CRITICAL POC
10.0 Oct 02

Loom for AWS deployments running a version before 1.6.1 with no identity provider configured expose an unauthenticated s

CVE-2026-27702 CRITICAL POC
9.9 Feb 25

Unauthorized data access in Budibase low-code platform before 3.30.4 allows unauthenticated users to manipulate internal

CVE-2025-64709 CRITICAL POC
9.9 Nov 13

Authenticated Typebot users can abuse the webhook block (HTTP Request component) in versions prior to 3.13.1 to issue ar

CVE-2020-37153 CRITICAL POC
9.8 Feb 11

Multiple vulnerabilities in ASTPP 4.0.1 including XSS and command injection in SIP device configuration and plugin manag

CVE-2026-87911 CRITICAL POC
9.0 Sep 09

Remote code execution in Amazon awslabs postgres-mcp-server before 1.1.7 lets remote actors run arbitrary OS commands on

CVE-2026-11393 HIGH POC
8.8 Jun 08

Remote code execution in AWS AgentCore CLI before v0.14.2 allows authenticated attackers to inject Python code via craft

CVE-2026-18140 HIGH POC
8.7 Jul 30

Denial of service in the aws-smithy-json Rust runtime crate (versions before 0.62.7) lets remote unauthenticated attacke

CVE-2026-16756 HIGH POC
8.7 Jul 23

Denial of service in Amazon's aws-smithy-http-server Rust crate (versions prior to 0.66.5) lets remote unauthenticated a

CVE-2026-16796 HIGH POC
8.4 Jul 23

Argument injection in the AWS Bedrock AgentCore Python SDK before 1.18.1 lets a remote authenticated user run arbitrary

CVE-2026-9133 HIGH POC
8.3 May 20

Arbitrary file read in amazon-mq rabbitmq-aws before 0.2.1 allows authenticated remote users to read any file accessible

CVE-2026-13763 HIGH POC
7.9 Jun 29

WAF managed-rule body inspection on AWS Application Load Balancer (ALB) can be bypassed by remote actors who fragment an

CVE-2026-13762 HIGH POC
7.9 Jun 29

WAF inspection bypass in Amazon CloudFront (with AWS WAF enabled) lets remote actors smuggle malicious request bodies pa

This site uses cookies essential for authentication and security. No tracking or analytics cookies are used. Privacy Policy