Significant-Gravitas AutoGPT
AutoGPT's API Keys and Secrets Logged in Plaintext in Stagehand Integration Blocks
AutoGPT is a platform that allows users to create, deploy, and manage continuous artificial intelligence agents that automate complex workflows. Prior to autogpt-platform-beta-v0.6.46, the AutoGPT platform's Stagehand integration blocks log API keys and authentication secrets in plaintext using logger.info() statements. This occurs in three separate block implementations (StagehandObserveBlock, StagehandActBlock, and StagehandExtractBlock) where the code explicitly calls api_key.get_secret_value() and logs the result. This issue has been patched in autogpt-platform-beta-v0.6.46.
Affected versions and mitigation
From the CVE record and the vendor's references.Affected
- < autogpt-platform-beta-v0.6.46
Mitigation
The record names no fix. Check the vendor advisory below.
References
Timeline
Days relative to CVE publicationFirst public PoC 79 days after publication; not in CISA KEV. First PoC: sivaadityacoder/CVE-2026-22038 (created 25 Apr 2026, found by us: backfilled).
PoC repositories
Found on GitHub.1 PoCs
PoCs 1
| Repository | Created | Stars | Assessment |
|---|---|---|---|
| sivaadityacoder/CVE-2026-22038 | 25 Apr 2026 +79 d | 0 | PoC via PoC Index |
Only PoCs count toward the first-PoC date and the characteristics below. Do not run PoC code from unknown sources.
PoC characteristics
What to hunt for in your logs · from 1 PoC repoPOST /v1/chat/completions
- Key headers
Authorization: Bearer sk-proj-abc123xyz789...- Body markers
"model":"gpt-4""messages""role":"user""content":"test"inapplication/json- Seen in
- 1 repo: sivaadityacoder/CVE-2026-22038
- Evidence
- sivaadityacoder/CVE-2026-22038/README.md:102
Regex
Generated from PoC characteristics; not yet reviewed. Matches the request path and query.
(?i)^/v1/chat/completions(?:[?#].*)?$
Suricata
Template generated from PoC characteristics. Review before deploying.
alert http any any -> $HOME_NET any (msg:"GG PoC-Monitor CVE-2026-22038 Significant-Gravitas AutoGPT exploit attempt"; flow:established,to_server; http.method; content:"POST"; http.uri; content:"/v1/chat/completions"; http.header; content:"Authorization: Bearer sk-proj-abc123xyz789..."; nocase; http.request_body; content:"|22|model|22|:|22|gpt-4|22|"; http.request_body; content:"|22|messages|22|"; http.request_body; content:"|22|role|22|:|22|user|22|"; http.request_body; content:"|22|content|22|:|22|test|22|"; reference:url,poc.griffinguard.io/cve/CVE-2026-22038; classtype:web-application-attack; sid:9029855; rev:1;)
Sigma
Template generated from PoC characteristics. Review before deploying.
title: Significant-Gravitas AutoGPT exploit attempt (CVE-2026-22038)
id: gg-pocmon-cve-2026-22038-9029855
status: experimental
description: Request pattern taken from public PoCs for CVE-2026-22038. Generated, not reviewed.
references:
- https://poc.griffinguard.io/cve/CVE-2026-22038
logsource:
category: webserver
detection:
selection:
cs-method: POST
cs-uri-stem|contains|all:
- '/v1/chat/completions'
condition: selection
falsepositives:
- Authorized vulnerability scanning
level: high
tags:
- attack.initial_access
- attack.t1190
- cve.2026.22038