GHSA-4X48-CGF9-Q33F
Vulnerability from github – Published: 2026-04-14 23:22 – Updated: 2026-04-14 23:22Summary
The conditions filter webhook at libs/application-generic/src/usecases/conditions-filter/conditions-filter.usecase.ts line 261 sends POST requests to user-configured URLs using raw axios.post() with no SSRF validation. The HTTP Request workflow step in the same codebase correctly uses validateUrlSsrf() which blocks private IP ranges. The conditions webhook was not included in this protection.
Root Cause
conditions-filter.usecase.ts line 261:
return await axios.post(child.webhookUrl, payload, config).then((response) => {
return response.data as Record<string, unknown>;
});
No call to validateUrlSsrf(). The webhookUrl comes from the workflow condition configuration with zero validation.
Protected Code (for contrast)
execute-http-request-step.usecase.ts line 130:
const ssrfValidationError = await validateUrlSsrf(url);
if (ssrfValidationError) {
// blocked
}
This function resolves DNS and checks against private ranges (127.0.0.0/8, 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16, 169.254.0.0/16). It exists in the codebase but is not applied to the conditions webhook path.
Proof of Concept
- Create a workflow with a condition step
- Configure the condition's webhook URL to
http://169.254.169.254/latest/meta-data/iam/security-credentials/ - Trigger the workflow by sending a notification event
- The worker evaluates the condition and calls
axios.post()to the metadata endpoint - The response data is stored in execution details and accessible via the execution details API
Impact
Full-read SSRF. The response body is returned as Record<string, unknown> for condition evaluation and stored in the execution details raw field. The GET /execution-details API returns this data.
The POST method limits some metadata endpoints (GCP requires GET, Azure requires GET), but AWS IMDSv1 accepts POST and returns credentials. Internal services accepting POST are also reachable.
Suggested Fix
Extract validateUrlSsrf() to a shared utility and call it before the axios.post in conditions-filter.usecase.ts:
const ssrfError = await validateUrlSsrf(child.webhookUrl);
if (ssrfError) {
throw new Error('Webhook URL blocked by SSRF protection');
}
return await axios.post(child.webhookUrl, payload, config)...
{
"affected": [
{
"package": {
"ecosystem": "npm",
"name": "@novu/api"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "3.15.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [],
"database_specific": {
"cwe_ids": [
"CWE-918"
],
"github_reviewed": true,
"github_reviewed_at": "2026-04-14T23:22:48Z",
"nvd_published_at": null,
"severity": "HIGH"
},
"details": "## Summary\n\nThe conditions filter webhook at `libs/application-generic/src/usecases/conditions-filter/conditions-filter.usecase.ts` line 261 sends POST requests to user-configured URLs using raw `axios.post()` with no SSRF validation. The HTTP Request workflow step in the same codebase correctly uses `validateUrlSsrf()` which blocks private IP ranges. The conditions webhook was not included in this protection.\n\n## Root Cause\n\n`conditions-filter.usecase.ts` line 261:\n```typescript\nreturn await axios.post(child.webhookUrl, payload, config).then((response) =\u003e {\n return response.data as Record\u003cstring, unknown\u003e;\n});\n```\n\nNo call to `validateUrlSsrf()`. The `webhookUrl` comes from the workflow condition configuration with zero validation.\n\n## Protected Code (for contrast)\n\n`execute-http-request-step.usecase.ts` line 130:\n```typescript\nconst ssrfValidationError = await validateUrlSsrf(url);\nif (ssrfValidationError) {\n // blocked\n}\n```\n\nThis function resolves DNS and checks against private ranges (127.0.0.0/8, 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16, 169.254.0.0/16). It exists in the codebase but is not applied to the conditions webhook path.\n\n## Proof of Concept\n\n1. Create a workflow with a condition step\n2. Configure the condition\u0027s webhook URL to `http://169.254.169.254/latest/meta-data/iam/security-credentials/`\n3. Trigger the workflow by sending a notification event\n4. The worker evaluates the condition and calls `axios.post()` to the metadata endpoint\n5. The response data is stored in execution details and accessible via the execution details API\n\n## Impact\n\nFull-read SSRF. The response body is returned as `Record\u003cstring, unknown\u003e` for condition evaluation and stored in the execution details `raw` field. The `GET /execution-details` API returns this data.\n\nThe POST method limits some metadata endpoints (GCP requires GET, Azure requires GET), but AWS IMDSv1 accepts POST and returns credentials. Internal services accepting POST are also reachable.\n\n## Suggested Fix\n\nExtract `validateUrlSsrf()` to a shared utility and call it before the axios.post in conditions-filter.usecase.ts:\n\n```typescript\nconst ssrfError = await validateUrlSsrf(child.webhookUrl);\nif (ssrfError) {\n throw new Error(\u0027Webhook URL blocked by SSRF protection\u0027);\n}\nreturn await axios.post(child.webhookUrl, payload, config)...\n```",
"id": "GHSA-4x48-cgf9-q33f",
"modified": "2026-04-14T23:22:48Z",
"published": "2026-04-14T23:22:48Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/novuhq/novu/security/advisories/GHSA-4x48-cgf9-q33f"
},
{
"type": "WEB",
"url": "https://github.com/novuhq/novu/commit/87d965eb88340ac7cd262dd52c8015acd092dc68"
},
{
"type": "PACKAGE",
"url": "https://github.com/novuhq/novu"
}
],
"schema_version": "1.4.0",
"severity": [],
"summary": "Novu has SSRF via conditions filter webhook bypasses validateUrlSsrf() protection"
}
Sightings
| Author | Source | Type | Date |
|---|
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.