← Vulnerability feed

Vulnerability record · CVE-2026-45019 · published 25 August 2026

CVE-2026-45019: Server-side request forgery (SSRF) vulnerability

Chainlit is a Python framework for building production-ready conversational AI applications. From 2.4.0rc0 until 2.12.0, Chainlit deployments with features.mcp.enabled set to true in .chainlit/config.toml expose the POST /mcp endpoint without requiring authentication. For sse and streamable-http transports, ConnectSseMCPRequest and ConnectStreamableHttpMCPRequest in backend/chainlit/types.py accept a user-controlled url and optional headers dictionary without scheme validation, private-address filtering, or an allowlist. The connect_mcp handler in backend/chainlit/server.py passes these values to sse_client() or streamablehttp_client(), allowing the Chainlit server to make blind outbound requests to arbitrary internal or external services, including cloud metadata endpoints, with attacker-controlled Authorization and Cookie headers. The SSE URL sink has existed since 2.4.0rc0, while attacker-controlled header forwarding and streamable-http support were added in 2.6.4. The response is consumed internally and not returned, but the attacker can issue state-changing authenticated requests, discover internal services, scan ports, and probe metadata endpoints. This issue is fixed in version 2.12.0.

7.2 CVSS 3.1 High EPSS 0.43% · top 64.8% CWE-918 · Server-side request forgery (SSRF) Deferred
7.2CVSS 3.1 base score
0.43%EPSS exploitation probability, 30 days
NoNot in CISA KEV
0Affected product versions listed by NVD
4References
9 Sep 2026Last modified by NVD

Description

Chainlit is a Python framework for building production-ready conversational AI applications. From 2.4.0rc0 until 2.12.0, Chainlit deployments with features.mcp.enabled set to true in .chainlit/config.toml expose the POST /mcp endpoint without requiring authentication. For sse and streamable-http transports, ConnectSseMCPRequest and ConnectStreamableHttpMCPRequest in backend/chainlit/types.py accept a user-controlled url and optional headers dictionary without scheme validation, private-address filtering, or an allowlist. The connect_mcp handler in backend/chainlit/server.py passes these values to sse_client() or streamablehttp_client(), allowing the Chainlit server to make blind outbound requests to arbitrary internal or external services, including cloud metadata endpoints, with attacker-controlled Authorization and Cookie headers. The SSE URL sink has existed since 2.4.0rc0, while attacker-controlled header forwarding and streamable-http support were added in 2.6.4. The response is consumed internally and not returned, but the attacker can issue state-changing authenticated requests, discover internal services, scan ports, and probe metadata endpoints. This issue is fixed in version 2.12.0.

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:N

References

LinkTags
https://github.com/Chainlit/chainlit/blob/2.12.0/docs/security-advisory-2026-mcp.md#spl-2026-002--ssrf-via-mcp-streamabl
https://github.com/Chainlit/chainlit/commit/0565fd0eccb915fce159929598b053ed79f6e0c9
https://github.com/Chainlit/chainlit/releases/tag/2.12.0
https://github.com/Chainlit/chainlit/security/advisories/GHSA-hvfh-5mj3-5f3j

Track CVE-2026-45019 inside VULONE

Watch it alongside the ransomware crews, C2 infrastructure and forum chatter that reference it, query it through the API and pull it into your SIEM over TAXII.

Start free Open in platform

Related vulnerabilities

Same products first, then exploited flaws of the same weakness class.

10.0CVE-2026-83548SonicWall SMA1000 pre-auth SSRF via alternate access pathThe SMA1000 Appliance Work Place interface exposes an unintended alternate access path that allows server-side request forgery before authentication.…KEVEPSS 8.8%analysed10.0CVE-2026-49869Kestra OSS auth bypass via path suffix match enables RCEKestra OSS AuthenticationFilter whitelists the public config endpoint using request.getPath().endsWith("/configs"), a suffix match instead of an exac…KEVEPSS 2.1%analysed9.3CVE-2026-64849MLflow unauthenticated webhook test endpoint SSRF via redirectMLflow before 3.15.0 validates the webhook URL only on the original request, while the delivery code follows redirects and re-resolves the hostname w…KEVEPSS 9.8%analysed10.0CVE-2026-15409SonicWall SMA1000 Work Place SSRF allows unauthenticated requestsThe SMA1000 Appliance Work Place interface contains a server-side request forgery flaw (CWE-918) that lets the appliance be induced to make requests …KEVEPSS 6.8%analysed8.6CVE-2026-20230Cisco Unified CM SSRF enables file write and root escalationCisco Unified Communications Manager and Unified CM SME fail to properly validate input for specific HTTP requests, allowing server-side request forg…KEVEPSS 88%analysed7.5CVE-2021-22054VMware Workspace ONE UEM console unauthenticated SSRFThe Workspace ONE UEM console contains a server-side request forgery flaw in versions 20.0.8 before 20.0.8.37, 20.11.0 before 20.11.0.40, 21.2.0 befo…KEVEPSS 100%analysed9.8CVE-2021-22175GitLab unauthenticated SSRF via internal webhook requestsGitLab is vulnerable to server-side request forgery when requests to the internal network for webhooks are enabled. The flaw affects all versions sta…KEVEPSS 53%analysed9.8CVE-2020-7796Zimbra Collaboration Suite WebEx zimlet SSRFZimbra Collaboration Suite before 8.8.15 Patch 7 is vulnerable to server-side request forgery when the WebEx zimlet is installed and its JSP is enabl…KEVEPSS 84%analysed

Source: NIST National Vulnerability Database (record CVE-2026-45019), CISA KEV, FIRST EPSS (scores of 2026-09-26). This page is refreshed as NVD updates the record.