# LucidRook Lua-Based Malware Toolkit Used by UAT-10362 in Targeted Espionage Against Taiwanese Civil Society

> Cisco Talos uncovered a new threat cluster tracked as UAT-10362 conducting spear-phishing campaigns against Taiwanese non-governmental organizations (NGOs) and universities. The campaigns deliver a novel three-component Lua-based malware toolkit — LucidPawn (dropper), LucidRook (Lua bytecode stager), and LucidKnight (Gmail-based reconnaissance tool) — representing new TTPs from a suspected China-nexus threat actor targeting Taiwan's civil society.

- **Published:** 2026-04-08T12:00:00Z
- **Last reviewed:** 2026-04-08T12:00:00Z
- **Canonical:** https://intel.threadlinqs.com/threat/TL-2026-0334
- **Card image:** https://intel.threadlinqs.com/og/threat/threat-intel-lucidrook-lua-toolkit-tl-2026-0334.webp?v=e0d7cdc7 (Threadlinqs threat intel: LucidRook Lua Taiwanese Civil Society: Detection & IOCs (TL-2026-0334) — High severity, 26 IOCs, 9 detection rules.)
- **ID:** TL-2026-0334
- **Severity:** HIGH
- **Category:** MALWARE
- **Status:** ACTIVE
- **Actor:** UAT-10362 (China)
- **Detections:** 9 (1 shown free below) · **IOCs:** 26 (3 shown free below)

## Description

Cisco Talos researchers identified a previously undocumented threat activity cluster designated UAT-10362 conducting targeted spear-phishing campaigns against Taiwanese non-governmental organizations (NGOs) and academic institutions. The campaigns deliver a novel malware toolkit collectively referred to as the Lucid family, consisting of three distinct components operating in a coordinated attack chain.

The initial infection vector is spear-phishing emails containing malicious attachments tailored to Taiwanese civil society targets. Upon execution, the LucidPawn dropper component establishes initial foothold on the victim system and deploys the primary payload. LucidRook, the core component of the toolkit, is a Lua bytecode stager that leverages the lightweight Lua scripting engine (MITRE ATT&CK T1059.011) to execute obfuscated bytecode payloads. By using compiled Lua bytecode rather than plaintext scripts, the malware evades static analysis and signature-based detection tools that lack Lua bytecode parsing capabilities.

The third component, LucidKnight, is a reconnaissance and data collection tool that abuses Gmail as a command-and-control channel. By tunneling C2 communications through legitimate Gmail infrastructure, the malware blends with normal email traffic and bypasses network security controls that would flag connections to suspicious domains. The toolkit also uses the domain digimg.store for DNS-based interaction callbacks, a technique commonly observed in post-exploitation frameworks for verifying command execution on target systems.

The two command-and-control IP addresses identified (1.34.253.131 and 59.124.71.242) both resolve to Chunghwa Telecom (HINET) address space in Taiwan, suggesting the threat actor either compromised legitimate Taiwanese infrastructure for use as relay nodes or deliberately selected Taiwanese hosting to blend with expected network traffic patterns from the target organizations.

UAT-10362 is assessed with moderate confidence to be a China-nexus threat actor based on victimology (Taiwanese NGOs and universities aligned with civil society and democracy advocacy), operational patterns consistent with Chinese cyber espionage campaigns, and infrastructure choices. The targeting of Taiwan's civil society sector is consistent with long-standing Chinese intelligence collection priorities, particularly monitoring organizations involved in cross-strait relations, democracy promotion, and academic research on sensitive geopolitical topics.

The use of Lua as the primary execution engine represents an evolution in tradecraft for China-nexus actors, who have historically favored compiled languages or common scripting engines. Lua's lightweight footprint, embeddability, and relative obscurity in enterprise security monitoring make it an attractive choice for evading endpoint detection and response (EDR) solutions that may not have robust Lua bytecode analysis capabilities.

## MITRE ATT&CK

- T1566 Phishing
- T1059 Command and Scripting Interpreter
- T1204 User Execution
- T1053 Scheduled Task/Job
- T1547 Boot or Logon Autostart Execution
- T1027 Obfuscated Files or Information
- T1036 Masquerading
- T1140 Deobfuscate/Decode Files or Information
- T1082 System Information Discovery
- T1033 System Owner/User Discovery
- T1057 Process Discovery
- T1083 File and Directory Discovery
- T1005 Data from Local System
- T1071 Application Layer Protocol
- T1102 Web Service
- T1105 Ingress Tool Transfer
- T1041 Exfiltration Over C2 Channel
- T1567 Exfiltration Over Web Service

## Free detection (Splunk SPL, Microsoft KQL, Sigma)

**LucidRook — C2 Communication via Gmail API Abuse and DNS Callbacks to digimg.store** (`TL-2026-0334-SPL-003`)

Detects command-and-control communication patterns associated with LucidKnight reconnaissance tool: anomalous Gmail SMTP/API traffic from non-browser processes, DNS queries to digimg.store callback infrastructure, and network connections to known UAT-10362 C2 IP addresses (1.34.253.131, 59.124.71.242). These C2 channels allow the threat actor to blend with legitimate traffic while maintaining persistent access.

- **Languages:** Splunk SPL, Microsoft KQL, Sigma
- **ATT&CK:** T1071.003, T1071.004, T1102, T1105
- **Data sources:** sysmon, firewall, dns_logs, proxy_logs, edr_network_events
- **Severity:** critical
- **Confidence:** high

### Splunk SPL

```spl
index=sysmon sourcetype="XmlWinEventLog:Microsoft-Windows-Sysmon/Operational" (EventCode=3 OR EventCode=22)
| search (
    (EventCode=3 (DestinationIp="1.34.253.131" OR DestinationIp="59.124.71.242"))
    OR (EventCode=3 DestinationPort=587 NOT Image="*\\chrome.exe" NOT Image="*\\firefox.exe" NOT Image="*\\msedge.exe" NOT Image="*\\outlook.exe" NOT Image="*\\thunderbird.exe")
    OR (EventCode=22 QueryName="*digimg.store*")
    OR (EventCode=22 QueryName="*digimg.store" QueryName="*.2fcc7078.*")
)
| eval c2_type=case(
    EventCode=3 AND (DestinationIp="1.34.253.131" OR DestinationIp="59.124.71.242"), "known_c2_ip",
    EventCode=3 AND DestinationPort=587, "smtp_non_browser",
    EventCode=22 AND match(QueryName, "digimg\.store"), "dns_callback"
)
| stats count min(_time) as first_seen max(_time) as last_seen values(Image) as processes values(DestinationIp) as dest_ips values(QueryName) as dns_queries values(DestinationPort) as dest_ports by Computer, c2_type, User
| table Computer, c2_type, processes, dest_ips, dns_queries, dest_ports, User, first_seen, last_seen, count
```

### Microsoft KQL

```kql
let c2_ips = DeviceNetworkEvents
| where Timestamp > ago(24h)
| where RemoteIP in ("1.34.253.131", "59.124.71.242")
| project Timestamp, DeviceName, AccountName=InitiatingProcessAccountName, RemoteIP, RemotePort, InitiatingProcessFileName, C2Type="KnownC2IP";
let smtp_abuse = DeviceNetworkEvents
| where Timestamp > ago(24h)
| where RemotePort == 587
| where InitiatingProcessFileName !in~ ("chrome.exe", "firefox.exe", "msedge.exe", "outlook.exe", "thunderbird.exe")
| project Timestamp, DeviceName, AccountName=InitiatingProcessAccountName, RemoteIP, RemotePort, InitiatingProcessFileName, C2Type="SMTP_NonBrowser";
let dns_callback = DeviceEvents
| where Timestamp > ago(24h)
| where ActionType == "DnsQueryResponse"
| where AdditionalFields has "digimg.store"
| project Timestamp, DeviceName, AccountName=InitiatingProcessAccountName, RemoteIP="", RemotePort=0, InitiatingProcessFileName, C2Type="DNS_Callback";
union c2_ips, smtp_abuse, dns_callback
| summarize Count=count(), FirstSeen=min(Timestamp), LastSeen=max(Timestamp) by DeviceName, AccountName, C2Type, InitiatingProcessFileName, RemoteIP
| project DeviceName, AccountName, C2Type, InitiatingProcessFileName, RemoteIP, FirstSeen, LastSeen, Count
```

### Sigma

```yaml
title: LucidRook C2 via Known UAT-10362 Infrastructure and DNS Callbacks
id: c2050f03-95c7-432d-967f-39925919fc61
status: experimental
description: Detects network connections to known UAT-10362 C2 IP addresses, DNS queries to digimg.store callback domain, and suspicious non-browser SMTP connections indicative of LucidKnight Gmail-based C2 communication.
references:
    - https://blog.talosintelligence.com/new-lua-based-malware-lucidrook/
    - https://otx.alienvault.com/indicator/domain/digimg.store
author: Threadlinqs Team
date: 2026-04-08
tags:
    - attack.command_and_control
    - attack.t1071.003
    - attack.t1071.004
    - attack.t1102
    - attack.t1105
logsource:
    category: network_connection
    product: windows
detection:
    selection_c2_ip:
        DestinationIp:
            - '1.34.253.131'
            - '59.124.71.242'
    selection_dns_callback:
        DestinationHostname|endswith: '.digimg.store'
    selection_smtp_nonbrowser:
        DestinationPort: 587
        Image|endswith:
            - '\lua.exe'
            - '\luajit.exe'
            - '\python.exe'
            - '\powershell.exe'
            - '\cmd.exe'
    condition: selection_c2_ip or selection_dns_callback or selection_smtp_nonbrowser
falsepositives:
    - Legitimate connections to Chunghwa Telecom IP ranges for non-malicious purposes
    - Developer tools that query DNS callback services for testing
    - Custom mail-sending scripts using SMTP on port 587
level: critical
```

**Known false positives:**

- Legitimate connections to Chunghwa Telecom IP ranges for non-malicious purposes
- Developer tools that use DNS callback services like interact.sh for testing
- Custom mail-sending scripts using SMTP port 587 for authorized automation

**Tuning notes:** Remove C2 IP indicators after confirmed takedown. Add additional DNS callback domains (interact.sh, dnslog.cn) if broader coverage desired. Whitelist authorized SMTP-sending applications by process hash.

### 8 more detection rules (Splunk SPL, Microsoft KQL, Sigma), unlock with TL Creds

- LucidRook — Known Malware Hash Execution or Lua Bytecode File Creation (locked: unlock with TL Creds)
- LucidRook — System Discovery and Reconnaissance Burst from Lua or Scripting Engine (locked: unlock with TL Creds)
- LucidRook — Data Exfiltration via Gmail Web Service and Suspicious Outbound Transfers (locked: unlock with TL Creds)
- LucidRook — Obfuscated Lua Bytecode File Dropped in User Temp or AppData (locked: unlock with TL Creds)
- LucidRook — Lua Runtime Masquerading as System Service or Legitimate Application (locked: unlock with TL Creds)
- LucidRook — DNS Interaction Callback to [indicator] C2 Infrastructure (locked: unlock with TL Creds)
- LucidRook — Lua Bytecode Stager Execution via LucidPawn Dropper (locked: unlock with TL Creds)
- LucidRook — Persistence via Scheduled Task or Registry Run Key for Lua Payload (locked: unlock with TL Creds)

Sign up free to unlock the detection rules: https://threadlinqs.com/?modal=signup

## Sample indicators (3 of 26)

Defanged, TLP:CLEAR.

- IP address: `1.34.253[.]131`
- File hash: `d49761cdbea170dd17255a958214db392dc7621198f95d5eb5749859c603100a`
- Email address: `fexopuboriw972[@]gmail[.]com`

Full set of 26 indicators unlocks with TL Creds: https://threadlinqs.com/tl-creds

## Sources

- [Cisco Talos: New Lua-based malware LucidRook observed in targeted attacks against Taiwanese organizations](https://blog.talosintelligence.com/new-lua-based-malware-lucidrook/)
- [Cisco-Talos IOCs: LucidRook indicators of compromise](https://github.com/Cisco-Talos/IOCs/tree/main/2026/04)
- [MITRE ATT&CK T1059.011: Command and Scripting Interpreter - Lua](https://attack.mitre.org/techniques/T1059/011/)
- [Picus Security: T1059.011 Lua in MITRE ATT&CK Explained](https://www.picussecurity.com/resource/blog/t1059-011-lua)
- [AlienVault OTX: digimg.store threat intelligence](https://otx.alienvault.com/indicator/domain/digimg.store)
- [Proofpoint: China-Aligned Espionage Actors Target Taiwan](https://www.proofpoint.com/us/blog/threat-insight/phish-china-aligned-espionage-actors-ramp-up-taiwan-semiconductor-targeting)
- [Darktrace: digimg.store observed in Ivanti EPMM exploitation campaigns](https://www.darktrace.com/blog/ivanti-under-siege-investigating-the-ivanti-endpoint-manager-mobile-vulnerabilities-cve-2025-4427-cve-2025-4428)

## Full data

The other 8 detection rules (Splunk SPL / Microsoft KQL / Sigma) unlock with TL Creds; a free account includes daily creds: https://threadlinqs.com/?modal=signup
Indicator values unlock with TL Creds, at a lower price on Red and Purple; AI agents can query rules and IOCs through the Threadlinqs MCP server (Purple): https://intel.threadlinqs.com/mcp

Canonical: https://intel.threadlinqs.com/threat/TL-2026-0334
