CVE-2026-24764
Prompt Injection via Slack Channel Metadata in OpenClaw AI Assistant
Publication date: 2026-02-19
Last updated on: 2026-02-19
Assigner: GitHub, Inc.
Description
Description
CVSS Scores
EPSS Scores
| Probability: | |
| Percentile: |
Meta Information
Affected Vendors & Products
| Vendor | Product | Version / Range |
|---|---|---|
| openclaw | openclaw | to 2026.2.3 (exc) |
Helpful Resources
Exploitability
| CWE ID | Description |
|---|---|
| CWE-74 | The product constructs all or part of a command, data structure, or record using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify how it is parsed or interpreted when it is sent to a downstream component. |
| CWE-94 | The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment. |
Attack-Flow Graph
AI Powered Q&A
Can you explain this vulnerability to me?
This vulnerability affects OpenClaw (formerly Clawdbot), a personal AI assistant that users run on their own devices. In versions 2026.2.2 and below, when the Slack integration is enabled, the metadata of Slack channels such as the topic or description can be incorporated into the AI model's system prompt.
Since prompt injection is a known risk for large language model (LLM)-driven systems, this issue increases the risk by allowing untrusted Slack channel metadata to be treated as higher-trust system input. This means that malicious or manipulated Slack channel metadata could influence the AI's behavior in unintended ways.
The vulnerability has been fixed in version 2026.2.3.
How can this vulnerability impact me? :
This vulnerability can impact users by increasing the risk of prompt injection attacks on the AI assistant. Because untrusted Slack channel metadata is treated as higher-trust input, an attacker could manipulate channel topics or descriptions to influence the AI's responses or behavior.
Such manipulation could lead to incorrect, misleading, or harmful outputs from the AI assistant, potentially compromising the integrity and reliability of the system.
However, the CVSS score of 3.7 indicates a low to medium severity, with limited impact on confidentiality and integrity, and no impact on availability.
How does this vulnerability affect compliance with common standards and regulations (like GDPR, HIPAA)?:
I don't know
How can this vulnerability be detected on my network or system? Can you suggest some commands?
I don't know
What immediate steps should I take to mitigate this vulnerability?
To mitigate this vulnerability, upgrade OpenClaw to version 2026.2.3 or later, where the issue has been fixed.
Additionally, consider disabling the Slack integration if upgrading immediately is not possible, to reduce the risk of prompt injection via Slack channel metadata.