PraisonAI versions before 1.6.58 contain an SSRF vulnerability due to inadequate validation of hostnames in SpiderTools._validate_url. The function _host_is_blocked checks literal host encodings but does not resolve DNS names before network fetches such as scrape_page, crawl, extract_links, extract_text, or URL-mention fetches. An attacker can supply a hostname that resolves to loopback, private, link-local, or cloud metadata IP addresses, bypassing the SSRF policy without requiring a rebinding race condition. This allows exposure of internal responses to the agent. The vulnerability is addressed in praisonaiagents 1.6.58.

An attacker with the ability to control hostnames used by PraisonAI can bypass SSRF protections and access internal network resources or cloud metadata endpoints. This can lead to exposure of sensitive internal data (confidentiality impact) and partial integrity loss. The CVSS score is 8.5 (high severity), indicating a significant security risk but no direct availability impact.

Upgrade PraisonAI to version 1.6.58 or later, where the vulnerability is fixed. No other mitigations are specified. Patch status is confirmed by the vendor advisory indicating the fix in 1.6.58.