Perimeter Exploitation and NTLM Redirection

High severity TLP:CLEAR Part 1 of 2 Research by Rapid7 Published 18 Nov 2026

Hypothesis

An attacker is exploiting vulnerabilities in perimeter applications (like Flowise or Citrix) to gain initial access or force NTLM authentication to a remote UNC path for credential theft.

Brief

Why now

Rapid7 recently published a Metasploit Wrapup detailing several new modules for perimeter and development devices. These updates include exploits for Flowise, Citrix NetScaler, and GitLab. Attackers frequently use these vulnerabilities to achieve remote code execution (RCE) or to force the appliance to authenticate to a remote UNC path. This redirection allows the adversary to capture NTLM hashes for credential theft and relay attacks.

How the Hunt Flows

The hunt begins by scoping the environment. The first query identifies every host running the software mentioned in the source report. It searches software inventory for packages including Flowise, GitLab, NetScaler, and specific WordPress plugins. This narrows the focus to potentially vulnerable surfaces.

Once scoped, the hunt executes two parallel queries to gather evidence. The first query scans HTTP logs for specific URI paths associated with the new Metasploit modules, such as password reset or API linting endpoints. The second query baselines outbound SMB traffic on port 445. It specifically looks for rare connections to external IP addresses that have occurred less than five times within the lookback period.

An analyst agent then evaluates the combined results. It looks for a temporal correlation between a hit on an exploit URI and a near-simultaneous outbound SMB connection. If both occur on the same host within a tight window, it indicates the server responded to an exploitation attempt by attempting to authenticate to an external resource. This correlation provides a high-confidence signal of successful exploitation.

What the Hunt Cannot See

This hunt has two primary blind spots. First, if the perimeter application uses TLS and the environment lacks traffic decryption or decrypted HTTP logs, URI-based patterns remain invisible. The hunt cannot inspect the encrypted payload to identify the exploit path. Second, if a firewall blocks outbound port 445 at the network edge, the connection attempt may not appear in the network connection logs. In this scenario, the exploit might have succeeded in triggering the request, but the secondary stage of credential theft failed and left no network record.

In this series

Steps

  1. Query, scoping

    Identify vulnerable perimeter software

    Scope the hunt to hosts running software mentioned in the Metasploit wrap-up.

    sql, reads hb_software_inventory
    SELECT device_hostname, package_name, package_versionFROM hb_software_inventoryWHERE (LOWER(package_name) LIKE '%flowise%' OR LOWER(package_name) LIKE '%gitlab%' OR LOWER(package_name) LIKE '%netscaler%' OR LOWER(package_name) LIKE '%secure%email%' OR LOWER(package_name) LIKE '%wordpress%' OR LOWER(package_name) LIKE '%openbullet%')

    What a hit looks like. A list of hosts running potentially vulnerable perimeter or development applications.

  2. Query, detection candidate

    Hunt for exploitation URI patterns

    Find HTTP requests hitting specific endpoints associated with recent RCE modules.

    sql, reads hb_http_activity
    SELECT device_hostname, url_path, http_method, src_endpoint_ip, timeFROM hb_http_activityWHERE (instr(',' || '{{target_uris}}' || ',', ',' || LOWER(url_path) || ',') > 0 OR LOWER(url_path) LIKE '%/wp-content/plugins/bug-library/%')  AND time >= datetime('now', '-{{lookback_days}} days')  AND ('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0)

    What a hit looks like. Requests to sensitive paths like /forgot-password in Flowise or CI lint in GitLab.

  3. Query, baseline

    Outbound SMB connection baseline

    Identify anomalous outbound SMB connections that may indicate NTLM hash disclosure.

    sql, reads hb_network_connection
    SELECT device_hostname, dst_endpoint_ip, COUNT(*) AS connections, MIN(time) AS first_seenFROM hb_network_connectionWHERE dst_endpoint_port = 445  AND direction = 'outbound'  AND time >= datetime('now', '-{{lookback_days}} days')  AND ('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0)GROUP BY device_hostname, dst_endpoint_ipHAVING connections < 5

    What a hit looks like. Rare outbound port 445 connections from perimeter servers to external IPs.

  4. Agent triage

    Weigh perimeter evidence

    Determine if the HTTP exploitation activity and SMB redirection are part of the same attack chain.

  5. Decision

    Decide on containment

    Route the hunt based on the agent's confidence in the exploitation verdict.

  6. Response action

    Isolate affected host

    Contain the breach and prevent credential relaying.

  7. Analyst task

    Analyst manual validation

    Verify the automated findings and analyze the SMB destination.

  8. Analyst task

    Close out hunt report

    Document the outcome and update software inventory requirements.

Coverage

2 of 5 stages covered

Stage and ATT&CK techniqueCoveredHow, or why not
Exploitation of Perimeter and Developer Tools
T1190
Yes vulnerable-host-inventory, perimeter-http-exploitation
NTLMv2 Hash Disclosure via UNC Path
T1557.001
Yes smb-redirection-attempts
Multi-Protocol Fetch Payload Execution
T1105 T1059.003
Out of scope Belongs to another part of the 'Metasploit Wrapup: Unauth RCE modules in Network Perimeter Devices and much more' series.
Linux ELF Executable Payload Injection
T1574
Out of scope Belongs to another part of the 'Metasploit Wrapup: Unauth RCE modules in Network Perimeter Devices and much more' series.
Docker Container Persistence and Host Access
T1610 T1543.003
Out of scope Belongs to another part of the 'Metasploit Wrapup: Unauth RCE modules in Network Perimeter Devices and much more' series.

Blind spots

  • Needs TLS inspection or decrypted HTTP logs. Attackers can hide exploit payloads in HTTPS traffic, making URI-based detection impossible without decryption. It would answer: What were the full URI and body of the request on encrypted channels?
  • Needs Firewall logs with reject actions. If port 445 is blocked at the perimeter, the network connection surface may not show the attempt, leaving the successful exploit (initial access) undetected if it relies on SMB for the second stage. It would answer: Were there failed SMB connection attempts to external IPs?

Parameters and data

Parameters

ParameterTypeDefaultWhat it is
lookback_daysnumber14Days of history to examine.
scope_hostslist[host]noneSpecific hosts to investigate; leave empty to hunt across the entire estate.
target_urislist[string]/forgot-password, /api/v1/lint, /wp-content/plugins/bug-library/URI paths associated with the latest Metasploit module releases.

Data sources

SourceCategoryTelemetry
Endpoint telemetry (hb_ surfaces)endpointendpoint
Network telemetrynetworknetwork
Web server / proxy logssiemnetwork

Source

An open hunt.md file. It runs anywhere that reads the format.

hunt.md
---
analysis: A simple detection rule on the URI or the SMB port lacks context. This hunt
  correlates the vulnerability state (vulnerable software present) with the behavior
  (URI hit) and the specific outcome (uncommon outbound SMB connection) to settle
  on a verdict.
blind_spots:
- id: tls-encryption-blind-spot
  question: What were the full URI and body of the request on encrypted channels?
  requires: TLS inspection or decrypted HTTP logs
  risk: Attackers can hide exploit payloads in HTTPS traffic, making URI-based detection
    impossible without decryption.
  stage: initial-access-perimeter-vulnerabilities
- id: smb-egress-filtering
  question: Were there failed SMB connection attempts to external IPs?
  requires: Firewall logs with reject actions
  risk: If port 445 is blocked at the perimeter, the network connection surface may
    not show the attempt, leaving the successful exploit (initial access) undetected
    if it relies on SMB for the second stage.
  stage: credential-access-ntlm-disclosure
coverage:
- stage: initial-access-perimeter-vulnerabilities
  status: covered
  steps:
  - vulnerable-host-inventory
  - perimeter-http-exploitation
- stage: credential-access-ntlm-disclosure
  status: covered
  steps:
  - smb-redirection-attempts
- reason: 'Belongs to another part of the ''Metasploit Wrapup: Unauth RCE modules
    in Network Perimeter Devices and much more'' series.'
  stage: execution-fetch-payload-delivery
  status: out_of_scope
- reason: 'Belongs to another part of the ''Metasploit Wrapup: Unauth RCE modules
    in Network Perimeter Devices and much more'' series.'
  stage: persistence-linux-elf-injection
  status: out_of_scope
- reason: 'Belongs to another part of the ''Metasploit Wrapup: Unauth RCE modules
    in Network Perimeter Devices and much more'' series.'
  stage: persistence-privileged-container-mount
  status: out_of_scope
guardrails:
  claims: no_unsupported
  evidence: citation_required
  missing_data: not_benign
  telemetry: untrusted
hunt:
  applicability: campaign-specific
  handoff: keep-as-periodic-hunt
  justification: Exploitation of perimeter devices leads to immediate unauthenticated
    network access or high-value credential theft; verifying the integrity of these
    public-facing services is a critical security obligation.
  methodology: model-assisted
  trigger: intel-report
hypothesis: An attacker is exploiting vulnerabilities in perimeter applications (like
  Flowise or Citrix) to gain initial access or force NTLM authentication to a remote
  UNC path for credential theft.
labels:
- hunt
- attack.t1190
- attack.t1557.001
- credential access
- execution
- initial access
- persistence
name: Perimeter Exploitation and NTLM Redirection
parameters:
  lookback_days:
    default: '14'
    description: Days of history to examine.
    type: number
  scope_hosts:
    default: []
    description: Specific hosts to investigate; leave empty to hunt across the entire
      estate.
    type: list[host]
  target_uris:
    default:
    - /forgot-password
    - /api/v1/lint
    - /wp-content/plugins/bug-library/
    description: URI paths associated with the latest Metasploit module releases.
    type: list[string]
provenance:
  authors:
  - name: Huntbase hunt generation
    org: huntbase.io
  generated:
    by: huntbase-hunt-generation
    from: https://www.rapid7.com/blog/post/pt-metasploit-wrapup-unauth-rce-modules-in-network-perimeter-devices-and-much-more
    gates:
    - dry-run
    - lint
    - critic
    model: hb_google/gemini-3-flash-preview
rationale: Start by identifying all servers running Flowise, GitLab, Citrix, and Cisco
  gateway software. Use the HTTP activity surface to look for URIs associated with
  Metasploit modules, specifically focusing on unauthenticated endpoints.
references:
- name: "Rapid7 \u2014 Metasploit Wrapup: Unauth RCE modules in Network Perimeter\
    \ Devices"
  url: https://www.rapid7.com/blog/post/pt-metasploit-wrapup-unauth-rce-modules-in-network-perimeter-devices-and-much-more
related:
- hunt: smb-egress-to-non-rfc1918
  reason: A broader hunt for any outbound SMB traffic regardless of perimeter application
    exploitation.
  relation: sibling
scenario:
  stages:
  - name: Exploitation of Perimeter and Developer Tools
    observables:
    - HTTP POST to /forgot-password in Flowise
    - Unauthenticated CI lint API calls to GitLab
    - PHP file uploads to /wp-content/plugins/bug-library/
    - Command injection patterns in Citrix NetScaler login logs
    - SQL injection patterns in Cisco Secure Email Gateway S/MIME handling
    - CVE-2025-58434
    - CVE-2026-88771
    - CVE-2026-97229
    - CVE-2026-76443
    - CVE-2024-5450
    slug: initial-access-perimeter-vulnerabilities
    tactic: initial-access
    techniques:
    - T1190
  - name: NTLMv2 Hash Disclosure via UNC Path
    observables:
    - OpenBullet2 forced SMB authentication via UNC path proxy source
    - Outbound SMB (port 445) connections to attacker-controlled infrastructure
    - CVE-2026-39908
    slug: credential-access-ntlm-disclosure
    tactic: credential-access
    techniques:
    - T1557.001
  - name: Multi-Protocol Fetch Payload Execution
    observables:
    - Fetch commands using FTP, HTTP, HTTPS, or TFTP to download LoongArch64 payloads
    - Execution of chmod on downloaded files
    - Reverse shell or bind shell activity (e.g., shell_reverse_tcp)
    slug: execution-fetch-payload-delivery
    tactic: execution
    techniques:
    - T1105
    - T1059.003
  - name: Linux ELF Executable Payload Injection
    observables:
    - Code cave and segment injection in Linux ELF files
    - Architecture-specific trampolines for x86, x64, or AArch64
    - Modified ELF binaries on disk
    slug: persistence-linux-elf-injection
    tactic: persistence
    techniques:
    - T1574
  - name: Docker Container Persistence and Host Access
    observables:
    - Creation of privileged Docker containers
    - 'Bind-mounting host filesystem (e.g., C: drive on Windows or root on Linux)
      to a container'
    - Docker Desktop creating persistent Alpine containers on Windows
    - Portainer API activity to create containers with host access
    slug: persistence-privileged-container-mount
    tactic: persistence
    techniques:
    - T1610
    - T1543.003
  summary: This campaign involves the use of newly released Metasploit modules targeting
    unauthenticated RCE and authentication bypass vulnerabilities in perimeter devices
    and developer tools like Citrix NetScaler, Cisco Secure Email Gateway, Flowise,
    and GitLab. After initial access, attackers leverage specialized Linux ELF injection
    techniques and privileged Docker container creation for persistent access and
    payload execution.
series:
  index: 1
  slug: metasploit-wrapup-unauth-rce-modules-in-network-perimeter-devices-and-much-more
  title: 'Metasploit Wrapup: Unauth RCE modules in Network Perimeter Devices and much
    more'
  total: 2
severity: high
targets:
  analyst:
    name: Tier-2 analyst
    role: analyst
  endpoint:
    category: endpoint
    name: Endpoint telemetry (hb_ surfaces)
    telemetry:
    - endpoint
  hunter:
    agent: true
    name: Hunt agent
  network:
    category: network
    name: Network telemetry
    telemetry:
    - network
  web:
    category: siem
    name: Web server / proxy logs
    telemetry:
    - network
tlp: clear
type: investigation
---


# Perimeter Exploitation and NTLM Redirection

This hunt targets recent Metasploit module releases for perimeter software. It first identifies hosts running vulnerable software, then concurrently examines HTTP logs for known exploitation paths and network logs for anomalous outbound SMB traffic (port 445). An agent evaluates whether these signals coincide, suggesting a successful exploit that redirected authentication to an attacker-controlled listener.

## vulnerable-host-inventory
<!-- Identify vulnerable perimeter software -->
Scope the hunt to hosts running software mentioned in the Metasploit wrap-up.

```sqlite target=endpoint role=scoping
~~~yaml
expected: A list of hosts running potentially vulnerable perimeter or development
  applications.
reads:
- device_hostname
- package_name
- package_version
silence: not_evidence_of_absence
source: hb_software_inventory
verified: dry-run
verified_at: '2026-11-18'
~~~
SELECT device_hostname, package_name, package_version FROM hb_software_inventory WHERE (LOWER(package_name) LIKE '%flowise%' OR LOWER(package_name) LIKE '%gitlab%' OR LOWER(package_name) LIKE '%netscaler%' OR LOWER(package_name) LIKE '%secure%email%' OR LOWER(package_name) LIKE '%wordpress%' OR LOWER(package_name) LIKE '%openbullet%')
```

## parallel-evidence-gathering
<!-- Parallel evidence gathering -->
parallel:
- → perimeter-http-exploitation
- → smb-redirection-attempts
join: → incident-triage

## perimeter-http-exploitation
<!-- Hunt for exploitation URI patterns -->
Find HTTP requests hitting specific endpoints associated with recent RCE modules.

```sqlite target=web role=detection-candidate params=(target_uris=target_uris, lookback_days=lookback_days, scope_hosts=scope_hosts)
~~~yaml
expected: Requests to sensitive paths like /forgot-password in Flowise or CI lint
  in GitLab.
reads:
- device_hostname
- url_path
- http_method
- src_endpoint_ip
- time
silence: not_evidence_of_absence
source: hb_http_activity
verified: dry-run
verified_at: '2026-11-18'
~~~
SELECT device_hostname, url_path, http_method, src_endpoint_ip, time FROM hb_http_activity WHERE (instr(',' || '{{target_uris}}' || ',', ',' || LOWER(url_path) || ',') > 0 OR LOWER(url_path) LIKE '%/wp-content/plugins/bug-library/%') AND time >= datetime('now', '-{{lookback_days}} days') AND ('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0)
```

## smb-redirection-attempts
<!-- Outbound SMB connection baseline -->
Identify anomalous outbound SMB connections that may indicate NTLM hash disclosure.

```sqlite target=network role=baseline params=(lookback_days=lookback_days, scope_hosts=scope_hosts)
~~~yaml
baseline:
  compare: first_seen
  window: '{{lookback_days}}d'
expected: Rare outbound port 445 connections from perimeter servers to external IPs.
prevalence:
  by: device_hostname
  key:
  - dst_endpoint_ip
  rare_below: 5
reads:
- device_hostname
- dst_endpoint_ip
- dst_endpoint_port
- direction
- time
silence: not_evidence_of_absence
source: hb_network_connection
verified: dry-run
verified_at: '2026-11-18'
~~~
SELECT device_hostname, dst_endpoint_ip, COUNT(*) AS connections, MIN(time) AS first_seen FROM hb_network_connection WHERE dst_endpoint_port = 445 AND direction = 'outbound' AND time >= datetime('now', '-{{lookback_days}} days') AND ('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0) GROUP BY device_hostname, dst_endpoint_ip HAVING connections < 5
```

## incident-triage
<!-- Weigh perimeter evidence -->
```agent target=hunter
cite: required
context:
- vulnerable-host-inventory
- perimeter-http-exploitation
- smb-redirection-attempts
max_iterations: 4
objective: Determine if the perimeter server was successfully exploited based on URI
  hits and subsequent outbound SMB activity.
success_criteria: A verdict of malicious if exploit URIs match outbound SMB redirections.
tools:
- endpoint
- network
- web
```

## remediation-decision
<!-- Decide on containment -->
if~: "the incident-triage verdict is malicious for at least one host" (confidence: high, judge=hunter)
then: → isolate-target
indeterminate: → manual-validation
unavailable: → manual-validation (blind_spot: tls-encryption-blind-spot)
else: → close-out-report

## isolate-target
<!-- Isolate affected host -->
```action target=endpoint
~~~yaml
approval: required
~~~
Isolate the endpoint to stop outbound SMB NTLM relaying and further RCE execution.
```
→ manual-validation

## manual-validation
<!-- Analyst manual validation -->
```manual target=analyst
Review the HTTP paths and SMB destination IP. Check if the SMB traffic follows the HTTP POST request immediately. Determine if the destination IP has a reputation for hosting NTLM relay listeners.
```
→ close-out-report

## close-out-report
<!-- Close out hunt report -->
```manual target=analyst
Summarize the findings. If Benign, record the URI hits as false positives for tuning. If Malicious, escalate to Incident Response.
```
→ end

Run this hunt against your own data.

In Huntbase, Scout runs every step across the tools you already connect, checks back through your history and returns a verdict with the evidence. The hunt.md file is open, so you can also take it anywhere that reads the format.

Drafted by machine (huntbase-hunt-generation, hb_google/gemini-3-flash-preview), checked by dry-run, lint, critic, then reviewed by a person.

Content pack

Sign up to open

Pack contents open in Huntbase.