CVE-2026-35665: 
OpenClaw (formerly Moltbot or Clawdbot) vulnerability analysis and mitigation

CVE-2026-35665 is a Denial of Service vulnerability in OpenClaw (npm package) caused by an incomplete fix for CVE-2026-32011, where the Feishu webhook handler retains overly permissive pre-authentication body parsing limits. Affected versions are all OpenClaw releases up to and including 2026.2.22 (for Node.js/npm); the issue was patched in version 2026.3.24. The vulnerability was published on April 10, 2026, with the advisory authored by steipete and assigned by VulnCheck. It carries a CVSS v3.1 base score of 5.3 (Medium) and a CVSS v4.0 base score of 6.9 (Medium) (GitHub Advisory).

The root cause is CWE-405 (Asymmetric Resource Consumption / Amplification) — specifically, the Feishu extension's webhook handler in extensions/feishu/src/monitor.ts was excluded from the CVE-2026-32011 patch that tightened body parsing limits across other webhook handlers. The vulnerable handler calls installRequestBodyLimitGuard with constants of 1MB body size and a 30-second timeout (FEISHU_WEBHOOK_MAX_BODY_BYTES = 1024 * 1024, FEISHU_WEBHOOK_BODY_TIMEOUT_MS = 30_000) before performing any signature verification. Because authentication only occurs after the full body is read, an unauthenticated attacker can send concurrent slow HTTP POST requests (a Slowloris-style slow-body attack) to /feishu/events, holding each connection open for up to 30 seconds and buffering up to 1MB per connection, while patched handlers use 64KB/5s limits (GitHub Advisory).

Successful exploitation results in a Denial of Service against any OpenClaw instance running the Feishu channel in webhook mode. With approximately 50 concurrent slow connections, an attacker can exhaust the server's connection handling capacity for 30-second waves, consume up to 50MB of memory per attack wave, and block all legitimate Feishu webhook deliveries — meaning bot messages will not be received. There is no confidentiality or integrity impact; the vulnerability is limited to availability (GitHub Advisory).

A proof-of-concept exploit is publicly available in the GitHub Security Advisory, including concrete curl commands and bash scripts that demonstrate the slow-body DoS technique against the vulnerable endpoint. No authentication, credentials, or special privileges are required — only network access to the publicly exposed Feishu webhook endpoint. There is no evidence of in-the-wild exploitation at this time, and the EPSS score is approximately 0.039% (very low probability of exploitation in the near term). The vulnerability is not listed in the CISA KEV catalog (GitHub Advisory).

1. Reconnaissance: Identify publicly accessible OpenClaw instances running the Feishu channel in webhook mode (versions ≤ 2026.2.22). The Feishu webhook endpoint (/feishu/events) must be internet-accessible for Feishu to deliver webhooks, making it a natural attack surface.
2. Craft slow HTTP POST requests: Prepare HTTP POST requests to /feishu/events with Content-Type: application/json and a declared Content-Length of up to 65536 bytes, but deliver the body at an extremely slow rate (e.g., 1 byte/second).
3. Send concurrent slow connections: Launch multiple simultaneous slow-body connections. Example bash command from the advisory:

for i in $(seq 1 10); do
  (echo -n 'A'; sleep 15) | \
  curl -s -o /dev/null -X POST http://<target>/feishu/events \
  -H "Content-Type: application/json" \
  -H "Content-Length: 65536" \
  --data-binary @- --max-time 35 &
done
wait

4. Hold connections open: Each connection is held open for up to 30 seconds (the server's timeout), consuming a connection slot and up to 1MB of buffer memory. With 50 concurrent connections, the server's connection pool is exhausted.
5. Block legitimate traffic: While attacker connections occupy server resources, legitimate Feishu webhook deliveries are queued or dropped, causing the bot to miss messages for the duration of the attack (GitHub Advisory).

• Network: High volume of concurrent, long-lived HTTP POST connections to the /feishu/events endpoint from one or more source IPs; connections that remain open for 15–30 seconds while transmitting minimal data (e.g., <100 bytes over 15+ seconds).
• Logs: Server access logs showing many simultaneous POST requests to /feishu/events with large declared Content-Length values but very slow data transfer rates; HTTP 408 (Request Timeout) responses being returned to slow-body connections.
• Process/Resource: Elevated memory consumption on the Node.js process (up to 50MB per attack wave of 50 connections); high number of open file descriptors or TCP connections associated with the OpenClaw process.
• Application Behavior: Legitimate Feishu webhook deliveries failing or timing out during attack windows; bot missing incoming messages from Feishu during periods of connection exhaustion (GitHub Advisory).

The primary remediation is to upgrade OpenClaw to version 2026.3.24 or later, which applies the corrected body parsing limits (64KB body size, 5-second timeout) to the Feishu webhook handler, consistent with the CVE-2026-32011 patch applied to other handlers. As a temporary workaround prior to patching, operators should implement rate limiting and connection timeout enforcement at the reverse proxy or load balancer layer in front of the /feishu/events endpoint, and restrict access to known Feishu IP ranges where possible. Additionally, monitoring for anomalous connection patterns to the Feishu webhook endpoint can help detect active exploitation attempts (GitHub Advisory).

The vulnerability was published by steipete via the GitHub Security Advisory program and assigned by VulnCheck. It was noted on CVE tracking platforms including VulnDB and CVEFeed shortly after disclosure. No significant broader media coverage or notable researcher commentary beyond the advisory itself has been identified at this time (GitHub Advisory).