Executive brief
OpenClaw, an automation tool, contains a security flaw in how it handles Feishu webhooks and card actions. Due to insecure default settings, the system may fail to verify the identity of incoming requests, allowing unauthorized users to bypass security checks. This could enable an attacker to remotely execute arbitrary commands, potentially leading to a full system compromise or unauthorized data access.
Technical details
OpenClaw versions prior to 2026.4.15 are vulnerable to an authentication bypass (CWE-287) and insecure default initialization (CWE-1188) within the Feishu extension. The vulnerability stems from a 'fail-open' logic where missing 'encryptKey' configurations or blank callback tokens are treated as valid, effectively bypassing signature verification and replay protection. A remote, unauthenticated attacker can exploit this by sending malformed webhook or card-action traffic to reach the command dispatch component. This allows for arbitrary command execution within the context of the application. The issue is resolved in version 2026.4.15, which enforces 'fail-closed' validation.
Affected products
- OpenClaw openclaw < 2026.4.15
Timeline
- 2026-04-14: patched: Fix commit c8003f1 pushed to repository.
- 2026-04-16: advisory: Original advisory GHSA-xh72-v6v9-mwhc published.
- 2026-05-06: disclosed: CVE-2026-44109 published.