Skip to main content

AutoGPT EUVDEUVD-2025-210282

| CVE-2025-32436 HIGH
Uncontrolled Resource Consumption (CWE-400)
2026-06-18 GitHub_M
7.1
CVSS 4.0 · Vendor: GitHub_M
View

Attack ChainAIDerived

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

  1. Recon
    Authenticate to AutoGPT instance
  2. Delivery
    Build workflow with StepThroughItemsBlock loop
  3. Exploit
    Chain AddAudioToVideoBlock and MediaDurationBlock
  4. Install
    Supply large URL list as iterator input
  5. C2
    Execute workflow to spawn unbounded screenshots
  6. Execute
    Working directory fills until disk exhausted
  7. Impact
    Service and co-located workloads denied

Vulnerability Assessment

Weakness
CWE
OWASP Top 10:2025
A06:2025 - Insecure Design
ASVS 5.0
V2 - Validation and Business Logic Resource consumption limits
MITRE ATT&CK

1 technique across 1 tactic: Impact

The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up.

Exploitation Conditions

Attacker requires an authenticated AutoGPT account with permission to create and execute workflows on a target instance running a version prior to 0.6.63 (PR:L confirms low-privilege auth is required). Exploitation specifically depends on access to the AddAudioToVideoBlock, StepThroughItemsBlock, and MediaDurationBlock components and the ability to supply a list of URLs for screenshotting; no user interaction by another party is needed (UI:N). Limiting factors: single-tenant self-hosted instances effectively require the operator to attack themselves, and any environment that already enforces disk quotas on the AutoGPT working directory will see workflow failure rather than host-wide DoS.

Impact

The CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:L) and base score of 7.1 paint a coherent picture: network-reachable, low attack complexity, requires low-privilege authentication, and impact is limited to availability of the vulnerable system with a minor subsequent-system availability impact (likely the host filesystem affecting co-located services). There is no confidentiality or integrity exposure and no remote code execution path. EPSS, KEV, and SSVC data were not provided in the input, but the absence of a CISA KEV listing and no referenced public PoC suggest this is not under active exploitation. The realistic risk is moderate and primarily relevant to multi-tenant or publicly accessible AutoGPT deployments where an authenticated low-trust user could deny service to other tenants; single-user self-hosted instances face minimal risk since the attacker would be the operator.

Exploit ScenarioDerived

An authenticated low-privilege user on a shared or public AutoGPT instance builds a workflow where StepThroughItemsBlock iterates over a large list of URLs, invoking MediaDurationBlock and AddAudioToVideoBlock on each item to screenshot pages and produce video output. Because the loop is unbounded and temporary media files are never cleaned up between iterations, the working directory fills until the host disk is exhausted, taking the AutoGPT service - and potentially other services on the same volume - offline. No public exploit code is referenced; exploitation requires only knowledge of which blocks to chain.

This Week Schedule remediation this week. Why: HIGH severity (CVSS 7.0-8.9).
Patch available 0.2% exploitation probability (EPSS)

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

Alert me

AnalysisAIv1 – Jun 18, 2026

Disk exhaustion denial-of-service in Significant Gravitas AutoGPT prior to version 0.6.63 allows authenticated platform users to crash the host by abusing the AddAudioToVideoBlock and StepThroughItemsBlock workflow components. By looping MediaDurationBlock through unbounded StepThroughItemsBlock iterations against URLs that trigger screenshots, attackers fill the working directory with undeleted temporary media files until disk space is exhausted. No public exploit identified at time of analysis, and EPSS data was not supplied.

RemediationAI

Vendor-released patch: upgrade AutoGPT to version 0.6.63 or later, which patches both the unbounded iteration in StepThroughItemsBlock and the missing cleanup logic in AddAudioToVideoBlock - see https://github.com/Significant-Gravitas/AutoGPT/security/advisories/GHSA-g26x-xwc5-7p44 for the upstream advisory. If immediate upgrade is not feasible, restrict the ability to create or run workflows containing AddAudioToVideoBlock and StepThroughItemsBlock to fully trusted accounts only, which has the side effect of breaking legitimate audio/video pipelines. Operators can additionally place AutoGPT's working directory on a dedicated, quota-enforced filesystem (e.g., a separate partition or XFS/ext4 project quota) so that exhaustion is contained and does not crash the host or co-located services, at the cost of failing workflows hitting the quota. Filesystem-level monitoring with alerts on rapid disk growth from the AutoGPT process can shorten detection time but does not prevent the DoS.

Affected VersionsENISA

AutoGPT < 0.6.63

Affected Packages

euvd autogpt fix: 0.6.63

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

AutoGPT is a workflow automation platform that chains together 'blocks' to build autonomous AI agent pipelines. The vulnerability sits in two cooperating blocks: AddAudioToVideoBlock, which downloads and stores video/audio assets in a temporary working directory without cleanup until the entire node graph completes, and StepThroughItemsBlock, which iterates over an input list with no upper bound on loop count. The root cause maps to CWE-400 (Uncontrolled Resource Consumption) - neither block enforces a quota on disk usage nor disposes of intermediate artifacts between iterations, so unbounded iteration multiplied by persistent temp files yields unbounded disk growth. The affected component is identified in CPE data as cpe:2.3:a:significant-gravitas:autogpt across all versions prior to the fix.

Know when Autogpt 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 (GitHub_M) PRIMARY 7.1 HIGH
CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:L
vuln.today AI 7.7 HIGH

Network-reachable workflow API with low-privilege account (PR:L), no user interaction, disk exhaustion can impact co-located services so scope changed (S:C) with high availability impact and no C/I impact.

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

Primary rating from Vendor (GitHub_M).

CVSS VectorVendor: GitHub_M

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

Lifecycle Timeline

1
Patch available
Jun 18, 2026 - 19:01 EUVD

DescriptionCVE.org

AutoGPT is a workflow automation platform for creating, deploying, and managing continuous artificial intelligence agents. Prior to 0.6.63, AddAudioToVideoBlock will download and store the video and audio in a temporary directory without deleting before all noded are done. StepThroughItemsBlock can be used to iterate MediaDurationBlock multiple times. StepThroughItemsBlock does not limit the number of loops. In addition, AddAudioToVideoBlock does not limit the amount of disk space consumed in the current working directory and does not delete the video after outputing the result. When a malicious user chooses to screen shot many web pages, the disk space will eventually run out, causing a DoS. Version 0.6.63 patches the issue.

Executive ViewAI

Why It Matters

AutoGPT versions prior to 0.6.63 contain a denial-of-service vulnerability that allows authenticated internal users to intentionally crash the platform by exhausting disk resources, directly impacting service availability and business continuity. Immediate access controls and disk monitoring are required to mitigate disruption risk.

Business Impact: High
Compensating Controls

No vendor-released patch identified at time of analysis; no public exploit identified at time of analysis. Interim mitigations: (1) Enforce role-based access restrictions on AddAudioToVideoBlock and StepThroughItemsBlock to limit exposure to trusted users only; (2) Deploy real-time disk space monitoring with automated alerts at 75-80% thresholds and automated remediation capability; (3) Implement automated temporary file cleanup with 24-hour retention in application working directory; (4) Enable comprehensive audit logging for workflow component access and execution patterns; (5) Establish documented incident response procedures with runbook for disk exhaustion detection and rapid remediation.

Threat Intelligence Overview

EPSS Score
0.2%
Top 84% exploitable
Social Media
1
Sightings
Data Quality
6 sources
Confidence: High
EUVD NVD CVE.org EPSS Vulnrichment CIRCL GitHub VulDB

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 0.2%
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
Scheduled
Exploitation
Poc
System Exposure· derived
Open
Automatable
No
Human Impact
High
Technical Impact
Partial

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

Denial of Service attacks render applications or systems unavailable by overwhelming resources or triggering failure conditions.

CVE-2026-55237 HIGH
8.8 Jun 18

DOM-based cross-site scripting in AutoGPT versions prior to 0.6.62 allows remote attackers to execute arbitrary JavaScri

CVE-2025-32392 HIGH
8.7 Jun 18

Resource exhaustion denial-of-service in Significant Gravitas AutoGPT versions prior to 0.6.63 allows remote unauthentic

CVE-2025-32437 HIGH
8.7 Jun 18

Denial of service in Significant-Gravitas AutoGPT prior to 0.6.63 allows remote attackers to exhaust host disk space by

CVE-2025-32424 HIGH
8.7 Jun 18

Denial of service in AutoGPT prior to 0.6.63 allows remote unauthenticated users to exhaust host disk space by chaining

CVE-2025-32422 HIGH
8.7 Jun 18

Denial of service in AutoGPT versions prior to 0.6.63 allows remote unauthenticated attackers to exhaust disk space on t

CVE-2026-56663 HIGH
8.5 Jun 26

Server-side request forgery in the AutoGPT Platform (versions prior to 0.6.52) lets an authenticated user abuse the Send

CVE-2026-72922 HIGH
8.2 Aug 11

Webhook signature verification bypass in AutoGPT Platform before 0.6.70 allows unauthenticated remote attackers to trigg

CVE-2026-33233 HIGH
7.6 May 19

Insecure deserialization in Significant-Gravitas AutoGPT platform versions 0.6.34 through 0.6.51 lets an attacker who ca

CVE-2026-33232 HIGH
7.5 May 19

Unauthenticated denial-of-service in AutoGPT Platform versions 0.4.2 through 0.6.51 allows remote attackers to exhaust s

CVE-2026-30950 HIGH
7.1 May 18

Authenticated session hijacking in Significant Gravitas AutoGPT versions 0.6.36 through 0.6.50 allows any logged-in user

CVE-2026-56823 MEDIUM
5.4 Jun 26

Improper access control in AutoGPT’s webhook ping endpoint allows any authenticated user to enumerate other users’ webho

CVE-2025-32394 MEDIUM
5.3 Jun 26

Memory exhaustion in AutoGPT's AITextSummarizerBlock allows authenticated users to trigger extreme resource amplificatio

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