GHSA-Q344-5V34-GM84
Vulnerability from github – Published: 2026-10-29 16:34 – Updated: 2026-10-29 16:34Impact
When @sylphx/pdf-reader-mcp is run as an HTTP server (MCP_TRANSPORT=http), it reads the MCP_API_KEY environment variable and logs API key authentication enabled (X-API-Key header) at startup, but it never validated the X-API-Key header on incoming requests. The key was read into a variable and never enforced on the request path.
As a result, the /mcp endpoint accepted unauthenticated requests. Combined with the default bind address of 0.0.0.0 (all interfaces), any client able to reach the port could invoke every PDF tool and read the contents of any PDF the server process could open anywhere on the host. URL-based sources also enable limited SSRF (private/loopback addresses are blocked by default unless --allow-private-ips is set).
The misleading startup log made the exposure silent: an operator who set MCP_API_KEY reasonably believed the server was protected when it was not.
This affects only the opt-in HTTP transport; the default stdio transport is unaffected.
Patches
Fixed in @sylphx/pdf-reader-mcp@3.0.1. The fix enforces the X-API-Key header with a constant-time comparison (rejecting missing/mismatched keys with HTTP 401), changes the default bind host to 127.0.0.1 (loopback), and emits a startup warning when binding a non-loopback host without an API key.
Workarounds
For deployments that cannot upgrade immediately:
- Bind to loopback only:
MCP_HTTP_HOST=127.0.0.1. - Place a reverse proxy in front that enforces authentication.
- Use the default
stdiotransport instead of HTTP. - Restrict filesystem reach with
--allow-dir/MCP_PDF_ALLOWED_DIRS.
Reported privately by novice-22 via responsible disclosure.
{
"affected": [
{
"package": {
"ecosystem": "npm",
"name": "@sylphx/pdf-reader-mcp"
},
"ranges": [
{
"events": [
{
"introduced": "2.2.0"
},
{
"fixed": "3.0.1"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-62264"
],
"database_specific": {
"cwe_ids": [
"CWE-200",
"CWE-287",
"CWE-306"
],
"github_reviewed": true,
"github_reviewed_at": "2026-10-29T16:34:35Z",
"nvd_published_at": null,
"severity": "HIGH"
},
"details": "### Impact\n\nWhen `@sylphx/pdf-reader-mcp` is run as an HTTP server (`MCP_TRANSPORT=http`), it reads the `MCP_API_KEY` environment variable and logs `API key authentication enabled (X-API-Key header)` at startup, but it never validated the `X-API-Key` header on incoming requests. The key was read into a variable and never enforced on the request path.\n\nAs a result, the `/mcp` endpoint accepted **unauthenticated** requests. Combined with the default bind address of `0.0.0.0` (all interfaces), any client able to reach the port could invoke every PDF tool and read the contents of any PDF the server process could open anywhere on the host. URL-based sources also enable limited SSRF (private/loopback addresses are blocked by default unless `--allow-private-ips` is set).\n\nThe misleading startup log made the exposure silent: an operator who set `MCP_API_KEY` reasonably believed the server was protected when it was not.\n\nThis affects only the opt-in HTTP transport; the default `stdio` transport is unaffected.\n\n### Patches\n\nFixed in `@sylphx/pdf-reader-mcp@3.0.1`. The fix enforces the `X-API-Key` header with a constant-time comparison (rejecting missing/mismatched keys with HTTP 401), changes the default bind host to `127.0.0.1` (loopback), and emits a startup warning when binding a non-loopback host without an API key.\n\n### Workarounds\n\nFor deployments that cannot upgrade immediately:\n\n- Bind to loopback only: `MCP_HTTP_HOST=127.0.0.1`.\n- Place a reverse proxy in front that enforces authentication.\n- Use the default `stdio` transport instead of HTTP.\n- Restrict filesystem reach with `--allow-dir` / `MCP_PDF_ALLOWED_DIRS`.\n\nReported privately by **novice-22** via responsible disclosure.",
"id": "GHSA-q344-5v34-gm84",
"modified": "2026-10-29T16:34:35Z",
"published": "2026-10-29T16:34:35Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/SylphxAI/pdf-reader-mcp/security/advisories/GHSA-q344-5v34-gm84"
},
{
"type": "WEB",
"url": "https://github.com/SylphxAI/anymd/pull/332"
},
{
"type": "WEB",
"url": "https://github.com/SylphxAI/anymd/commit/4c35f72862d729301e3912ab8c8317d2828cf4c2"
},
{
"type": "WEB",
"url": "https://github.com/SylphxAI/anymd/releases/tag/v3.0.1"
},
{
"type": "PACKAGE",
"url": "https://github.com/SylphxAI/pdf-reader-mcp"
}
],
"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": "@sylphx/pdf-reader-mcp: HTTP transport accepts unauthenticated requests despite a configured MCP_API_KEY"
}
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.
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
Loading timeline…
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
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.