Skip to content

SecurityChainlit

Fixed

ChainLeak: Chainlit file read and SSRF flaws (CVE-2026-22218, CVE-2026-22219) can expose secrets and cloud credentials

Chainlit at a glance

Two flaws in Chainlit before 2.9.4 let any logged-in user read files from the server or make it fetch internal URLs, including cloud metadata. Zafran Labs disclosed them as ChainLeak. Upgrade to 2.9.4.

Chainlit: status

FixedHigh
Affected
  • Chainlit versions before 2.9.4 (CVE-2026-22218)
  • Chainlit versions before 2.9.4 using the SQLAlchemy data layer (CVE-2026-22219)
Status as of
Discovered by
Gal Zaban and Ido Shani, Zafran Labs

What changed with Chainlit

On January 20, 2026, CVE-2026-22218 and CVE-2026-22219 were published for Chainlit, an open-source Python framework for building chat interfaces on top of LLM apps. Both live in the /project/element update flow and affect versions before 2.9.4. Zafran Labs researchers, who found them, call the pair ChainLeak.

CVE-2026-22218 (NVD 6.5, medium) lets an authenticated user submit a custom element with a file path, so the server copies any file it can read into the attacker's session for download. CVE-2026-22219 (NVD 7.7, high) is a server-side request forgery in deployments using the SQLAlchemy data layer: the server fetches an attacker-chosen URL, such as an internal service or cloud metadata endpoint, and stores the response.

Zafran reports that the file read can expose environment variables with API keys, the auth secret used to sign Chainlit sessions, source code, and cached prompts, and that on AWS instances still allowing IMDSv1 the SSRF can return IAM role credentials. Chainlit 2.9.4, released in December 2025, adds sanitization for custom element updates.

How to fix it: Chainlit

  1. 1. Update

    Upgrade Chainlit to 2.9.4 or later.

    nvd.nist.gov (opens in a new tab)

Advisories

Where to compare Chainlit

Independent leaderboards and official resources for comparing this model yourself. Rankings change often; each site shows its own current data.

Why Chainlit matters for businesses

Chainlit often fronts internal assistants and customer chatbots where any signed-up user counts as authenticated. A leaked auth secret or cloud credential turns one chat account into access to other users' sessions and the wider cloud account, so upgrade and enforce IMDSv2 on AWS hosts.

Sources for this story

  1. CVE-2026-22218 Detail (opens in a new tab)

    NIST National Vulnerability Database · nvd.nist.gov

    Official
  2. CVE-2026-22219 Detail (opens in a new tab)

    NIST National Vulnerability Database · nvd.nist.gov

    Official
  3. Chainlit 2.9.4 release (opens in a new tab)

    Chainlit on GitHub · github.com

    Official
  4. Press

Last verified: . IA Compa summarizes public announcements for business readers. We are not affiliated with the companies mentioned. Always confirm details, pricing, and availability with the official source.

  • #security
  • #cve
  • #chainlit
  • #ssrf
  • #chat-interfaces
  • #cloud-security
  • Generated cover for Grafana MCP server (mcp-grafana) by Grafana Labs: a code scan with flagged lines and a padlock, in Grafana Labs brand colors.

    Grafana LabsSecurity

    Grafana fixes critical SSRF in its MCP server that let callers reach internal services

    Grafana Labs published CVE-2026-19516 on August 11, 2026: in mcp-grafana, the X-Grafana-URL header and the grafana_api_request tool let a caller send requests to internal, loopback, and metadata endpoints and read the replies. Grafana rates it critical (CVSS 9.1); upgrade to 1.1.0 or later.

    Fixed
  • Generated cover for Azure MCP Server by Microsoft: a shield and padlock over a hexagon lattice, in Microsoft brand colors.

    MicrosoftSecurity

    Microsoft patches Azure MCP Server SSRF that could leak its managed identity token

    In its March 2026 security release, Microsoft fixed CVE-2026-26118, an SSRF in Azure MCP Server Tools rated Important (CVSS 8.8). Someone who can talk to an MCP-backed agent could swap a resource ID for a malicious URL and capture the server's managed identity token. Fixed in 1.0.2 and 2.0.0-beta.17.

    Fixed

Want this working in your business?

We turn new AI capabilities into agents, voice receptionists, and automations you can actually use.