CVE-2026-48039

Meta Ads MCP: Unauthenticated HTTP MCP Tool Execution Leaks Operator Meta Access Token

Published Aug 7, 2026 CVSS 3.1 · 9.1 Critical EPSS < 1% KEV · Not listed
Description
Meta Ads MCP is a Model Context Protocol (MCP) server that lets AI assistants run Meta Ads. Prior to version 1.0.109, `AuthInjectionMiddleware.dispatch()` at `http_auth_integration.py:272` unconditionally forwards unauthenticated Streamable HTTP requests to downstream MCP tool handlers without issuing a `401` response, allowing any network-reachable caller to invoke MCP tools without authentication. When no per-request credential is present, tool handlers fall back to the `META_ACCESS_TOKEN` environment variable, and when the downstream Meta Graph API call fails, `api.py:263–269` serialises the raw `httpx` request URL—including the operator's `access_token` as a query parameter—into the JSON-RPC response body, delivering the credential to the unauthenticated caller. Version 1.0.109 fixes the issue.

Analysis

AI-generated

Technical impact

Meta Ads MCP allows any network‑reachable caller to invoke MCP tools without providing authentication. In the affected versions the server falls back to an environment variable named META_ACCESS_TOKEN, and when a downstream Meta Graph API call fails, the raw HTTP request URL—including the access token as a query parameter—is serialized into the JSON‑RPC response body. The result is that an unauthenticated attacker can obtain a live Meta access token and potentially control the operator’s Meta Ads account. The flaw corresponds to CWE‑287: Authentication Bypass.

Affected Systems

The vulnerability affects Meta Ads MCP deployments running before version 1.0.109, as provided by pipeboard‑co.

Risk and Exploitability

With a CVSS score of 9.1 the vulnerability is considered critical. The EPSS score is not available, and the flaw is not listed in the CISA KEV catalog. The attack vector is any remote entity that can reach the MCP server’s HTTP interface, which is typically accessible from machines within the same network or from the public internet if not otherwise restricted.

Remediation

No solution or workaround provided in the CVE record.

OpenCVE Recommended Actions

  • Upgrade the Meta Ads MCP server to version 1.0.109 or later to apply the vendor patch.
  • Restrict network access to the MCP service so that only trusted hosts can reach its HTTP interface.
  • Ensure the META_ACCESS_TOKEN environment variable is stored securely and is not leaked in responses or logs.

Generated by OpenCVE AI on August 7, 2026 at 20:21 UTC.

Tracking

Sign in to view the affected projects.

Advisories

Source ID Title
Github GHSA Github GHSA GHSA-9gw6-46qc-99vr Meta Ads MCP: Unauthenticated HTTP MCP Tool Execution Leaks Operator Meta Access Token
History
Type Values Removed Values Added
First Time appeared Pipeboard-co
Pipeboard-co meta-ads-mcp
Vendors & Products Pipeboard-co
Pipeboard-co meta-ads-mcp
Type Values Removed Values Added
Metrics ssvc

{'options': {'Automatable': 'yes', 'Exploitation': 'poc', 'Technical Impact': 'total'}, 'version': '2.0.3'}

Type Values Removed Values Added
Description Meta Ads MCP is a Model Context Protocol (MCP) server that lets AI assistants run Meta Ads. Prior to version 1.0.109, `AuthInjectionMiddleware.dispatch()` at `http_auth_integration.py:272` unconditionally forwards unauthenticated Streamable HTTP requests to downstream MCP tool handlers without issuing a `401` response, allowing any network-reachable caller to invoke MCP tools without authentication. When no per-request credential is present, tool handlers fall back to the `META_ACCESS_TOKEN` environment variable, and when the downstream Meta Graph API call fails, `api.py:263–269` serialises the raw `httpx` request URL—including the operator's `access_token` as a query parameter—into the JSON-RPC response body, delivering the credential to the unauthenticated caller. Version 1.0.109 fixes the issue.
Title Meta Ads MCP: Unauthenticated HTTP MCP Tool Execution Leaks Operator Meta Access Token
Weaknesses CWE-287
References
Metrics cvssV3_1

{'score': 9.1, 'vector': 'CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:N'}

Weaknesses

Sources

Status
PUBLISHED
Assigner
GitHub_M
Published
Updated
Aug 7, 2026, 7:42 PM
Reserved
May 20, 2026, 6:15 PM
Status
Deferred
Published
Aug 7, 2026, 8:16 PM
Modified
Oct 23, 2026, 8:10 PM