PraisonAI Webhook Signature Verification Bypass

PraisonAI before 4.6.78 fails to verify Svix webhook signatures in AgentMail webhook mode, allowing unauthenticated attackers to forge message.received events. Attackers can send crafted JSON payloads to the webhook endpoint to invoke configured agents with arbitrary sender addresses and message content.

Casky was already ahead

This CVE exploits attack patterns that Casky's 283matched skills already investigate — long before this vulnerability was disclosed. Claude's reasoning model maps these techniques to MITRE ATT&CK, so practitioners who ran these skills have already seen the threat behaviour in their findings.

Analysis

PraisonAI versions before 4.6.78 contain a critical authentication weakness in AgentMail webhook mode where Svix webhook signatures are not properly validated. This allows unauthenticated attackers to forge message.received events and trigger configured agents with arbitrary sender addresses and message content. Organizations using PraisonAI for email-based agent automation are at direct risk of account compromise, data exfiltration, and unauthorized command execution through spoofed email messages that appear legitimate to deployed agents.

Casky's 283 matching skills map this vulnerability across TA0004 (Privilege Escalation) and TA0006 (Credential Access) techniques, enabling Claude AI to detect the attack chain through extended reasoning. Practitioners using Casky would identify suspicious patterns including: unauthenticated webhook requests bypassing signature verification (CWE-287 failure), anomalous agent invocations from unexpected senders, JSON payloads with mismatched or missing cryptographic signatures, and message.received events originating from non-whitelisted IPs or lacking proper HMAC validation headers. The platform would surface these signals alongside related techniques like T1078 (Valid Accounts) where attackers spoof trusted senders, and T1566 (Phishing) patterns where forged messages trigger unintended agent actions—allowing teams to correlate webhook anomalies with downstream agent behavior for rapid incident response.

Live Threat IntelligenceComing soon

Casky Risk Score

—

EPSS Probability

—

Shodan Exposure

—

GreyNoise Signal

—

Composite risk scoring from EPSS, CISA KEV, Shodan, and GreyNoise — 21 security APIs correlated into a single Casky Risk Score. Coming in Casky Pro. Join early access →

283 Casky skills cover this attack pattern

These skills use Claude AI's reasoning model to surface findings in the same attack categories as CVE-2026-61436.

abusing-dpapi-for-credential-access

red teaming · high

Run

abusing-shadow-credentials-for-privesc

red teaming · high

Run

Access with Stolen Session Cookie

identity access management · low

Run

MITRE ATT&CK Techniques

TA0004 TA0006

Investigate the attack patterns behind CVE-2026-61436

Casky has 283 skills that investigate the attack patterns behind CVE-2026-61436. Run one and get CVSS-scored findings in 3 minutes.

Run the skill that detects this →

X Instagram LinkedIn