Detection queries Reported campaignDeveloper workflows · Agents & MCP · Windows · macOS

Apex Accelerator / REF8801 backdoored AI CLI and MCP packages

A fake Web3 accelerator's npm packages route AI traffic to an attacker proxy, register rogue MCP servers in AI IDEs, and drop stealers on Windows and macOS.

What happened

Fake Web3 accelerator hands founders free Claude/Gemini credits and npm packages (@apexacc/cli, @apexacc/audit). CLI is Oh My Pi with backdoored bootstrap: OPENROUTER_BASE_URL to attacker LLM proxy, rogue MCP in ~/.apex/agent/mcp.json, yolo approvals. Audit MCP returns fabricated findings and registers into Claude Desktop/Cursor/Windsurf. Windows chain: wscript→PowerShell→stealer; macOS CloudSyncD.

Read the original source

How it happened

Select a stage to explore its defenses MagicSword policy opportunity
Detect this stage
Detect this stage
Detect this stage
MagicSword Detect this stage
Select a stage · Drag to explore · + / − to zoom · Dashed edges are loopsSelect a stage · Arrows show reported connectionsReported sequence
Read the flow as text & sources

A catalog reconstruction of the supplied source account. Unknown steps remain explicit; an edge describes the reported relationship between steps.

  1. 01 · Fake accelerator recruits founders

    Free AI credits entice Web3 founders to install npm packages.

    Hunt: An impersonating message, ad or website leads to a suspicious download, consent request, pasted command or payment instruction.

    Polished text, AI branding and a novel domain do not prove maliciousness or AI generation. Follow-on evidence matters.

    Source for step 1
  2. 02 · Bootstrap changes agent settings

    The packages introduce an attacker LLM proxy, rogue MCP configuration and permissive approvals.

    Hunt: A repository, dependency or agent changes MCP commands, automatic-task settings, tool approval or model-routing configuration before new execution.

    A file-write event alone cannot establish what changed. Legitimate setup also changes these files; obtain a diff and authorization context.

    Source for step 2
  3. 03 · Rogue tools influence the agent

    The audit MCP fabricates findings and registers with desktop coding agents.

    Hunt: Untrusted content or a changed tool definition precedes a file read, command, or remote call outside the user's intended task.

    Process logs cannot recover model intent. Remote connectors and in-process interpreters may produce no local child process. Prompt content alone is not a reliable signature.

    Source for step 3
  4. 04 · Platform-specific malware

    Reported Windows and macOS branches deliver a stealer or CloudSyncD.

    Hunt: A script host or shell runs encoded content, a remote URL, or inline script syntax, then downloads or launches a payload.

    Encoded commands and remote administration can be legitimate. This does not cover every script host, in-process execution, or non-Windows branch.

    MagicSword policy opportunity · The Windows wscript → PowerShell branch only. This does not cover the separate macOS CloudSyncD branch.

    Source for step 4

Why it matters

Reported entry point: Social engineering via fake accelerator / free AI credits; npm install of @apexacc/cli or @apexacc/audit. Reported impact: Reported Windows and macOS branches deliver a stealer or CloudSyncD; reported targets include Cryptocurrency / Web3 founders and startups.

Check these source-linked conditions against your setup.

CursorClaude / Claude Desktopnpm / PyPI dependenciesWindowsmacOS

The campaign registers rogue MCP configurations in AI clients after an attacker-supplied npm bootstrap runs.

Conditions to check
  • Confirm installation of the accelerator-supplied packages or the rogue configuration; ordinary AI-client use is not the entry point.
WindowsmacOSMCP connection required

An untrusted npm bootstrap can add rogue MCP servers and alter agent approval settings.

Conditions to check
  • The source describes a specific accelerator / npm package lure; check whether those packages or configurations are present.

Source-linked applicability guidance. Confirm current scope in the original report.

Read the original report

What you can do

Protection & prevention

Start here

Review newly added MCP servers, CLI bootstrap settings and the provenance of accelerator-supplied npm packages.

Block apexaccs.org / apexfdn.xyz / payload domains; audit MCP configs and OPENROUTER_BASE_URL; rotate AI keys; prevent wscript (Windows stealer chain depends on it)

Catalog defensive guidance based on the reported behavior. Confirm scope and current mitigations in the original source.

Review the original guidance LOLSkills
Explore scoped MagicSword policy options

Requires an enforced Windows policy covering the actual executable. Audit mode records events without blocking.

Restrict script / execution utilities

Use MagicSword to review and enforce a policy restricting unneeded Windows script hosts, shells or installer utilities identified in this branch.

Applies to: The Windows wscript → PowerShell branch only. This does not cover the separate macOS CloudSyncD branch.

Policy opportunity, not tested incident coverage. Requires enrolled Windows devices, a policy covering the actual executable, and enforcement. Audit mode records decisions without blocking; allowed tools and in-process / cloud actions can remain reachable.

Explore the policy control View mapped attack stage
Investigate each phase

Detection queries

Draft query starters grouped by the attack stages where they apply.

Download stage pack

Start with the logs you collect. Each hunt identifies its required visibility, field mapping and limits. Queries are catalog-authored drafts using reusable behavior profiles; test with known activity and your normal baseline before alerting.

01Fake accelerator recruits foundersLure → user action / deliveryCustom schema · Draft

Free AI credits entice Web3 founders to install npm packages.

What to look for

An impersonating message, ad or website leads to a suspicious download, consent request, pasted command or payment instruction.

Join the accelerator invitation to the package-install request; baseline legitimate recruitment and grants.

Visibility: Defender-side logs where collected

Required logs

  • Email / secure-web-gateway records
  • Browser redirects, downloads and URL reputation
  • User reports and downstream execution / identity events

Coverage limit. Polished text, AI branding and a novel domain do not prove maliciousness or AI generation. Follow-on evidence matters.

Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.

Sigma
Draft SigmaDownload Sigma
title: Hunt - Lure → user action / delivery
id: e0056b49-9905-485b-ac12-2aa076a7bc82
status: experimental
description: An impersonating message, ad or website leads to a suspicious
  download, consent request, pasted command or payment instruction.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
  - Authorized activity matching this broad hunt; baseline and investigate
    before alerting.
level: low
logsource:
  category: application
  product: openattacks_audit
  definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
    Target and CorrelationId from the required source logs first.
detection:
  selection:
    EventType:
      - message.link.click
      - web.download
  condition: selection
KQL
Draft KQLDownload KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("message.link.click", "web.download")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated desc
SPL
Draft SPLDownload SPL
sourcetype="openattacks:audit"
EventType IN ("message.link.click", "web.download")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _time
Investigation steps
  1. Reconstruct redirect and delivery provenance using original report indicators, without visiting payload URLs.
  2. Correlate the click with the actual download, consent or execution; verify the sender through an independent channel.
02Bootstrap changes agent settingsAgent trust / configuration changesCustom schema · Draft

The packages introduce an attacker LLM proxy, rogue MCP configuration and permissive approvals.

What to look for

A repository, dependency or agent changes MCP commands, automatic-task settings, tool approval or model-routing configuration before new execution.

Visibility: Defender-side logs where collected

Required logs

  • File changes and before/after content for MCP / IDE configuration
  • Repository audit / pull-request history
  • Agent startup and tool-invocation logs

Coverage limit. A file-write event alone cannot establish what changed. Legitimate setup also changes these files; obtain a diff and authorization context.

Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.

Sigma
Draft SigmaDownload Sigma
title: Hunt - Agent trust / configuration changes
id: 2dfc8034-d66b-42a7-a699-a76b2ef872ac
status: experimental
description: A repository, dependency or agent changes MCP commands,
  automatic-task settings, tool approval or model-routing configuration before
  new execution.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
  - Authorized activity matching this broad hunt; baseline and investigate
    before alerting.
level: low
logsource:
  category: application
  product: openattacks_audit
  definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
    Target and CorrelationId from the required source logs first.
detection:
  selection:
    EventType:
      - agent.config.write
      - agent.approval.change
  condition: selection
KQL
Draft KQLDownload KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("agent.config.write", "agent.approval.change")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated desc
SPL
Draft SPLDownload SPL
sourcetype="openattacks:audit"
EventType IN ("agent.config.write", "agent.approval.change")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _time
Investigation steps
  1. Diff command, args, environment, origin, autoApprove and task settings against the previously approved version.
  2. Correlate the author and change with the next agent launch; inspect path canonicalization and trailing-dot variants where relevant.
03Rogue tools influence the agentUnexpected agent tool useCustom schema · Draft

The audit MCP fabricates findings and registers with desktop coding agents.

What to look for

Untrusted content or a changed tool definition precedes a file read, command, or remote call outside the user's intended task.

Visibility: Defender-side logs where collected

Required logs

  • Agent tool-call audit records with session / tool / destination
  • Tool-definition and approval changes
  • Endpoint process, file-access and network evidence where available

Coverage limit. Process logs cannot recover model intent. Remote connectors and in-process interpreters may produce no local child process. Prompt content alone is not a reliable signature.

Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.

Sigma
Draft SigmaDownload Sigma
title: Hunt - Unexpected agent tool use
id: c709e384-0e11-479e-ab4d-3a06e2431e4b
status: experimental
description: Untrusted content or a changed tool definition precedes a file
  read, command, or remote call outside the user's intended task.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
  - Authorized activity matching this broad hunt; baseline and investigate
    before alerting.
level: low
logsource:
  category: application
  product: openattacks_audit
  definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
    Target and CorrelationId from the required source logs first.
detection:
  selection:
    EventType:
      - agent.tool.call
      - agent.tool.definition.change
  condition: selection
KQL
Draft KQLDownload KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("agent.tool.call", "agent.tool.definition.change")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated desc
SPL
Draft SPLDownload SPL
sourcetype="openattacks:audit"
EventType IN ("agent.tool.call", "agent.tool.definition.change")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _time
Investigation steps
  1. Join input provenance, tool descriptions and calls by session ID and compare the action to the approved task.
  2. Inspect child processes and destinations; record whether execution was in-process, local, or through a remote SaaS connector.
04Platform-specific malwareSuspicious Windows script executionNative logs · Draft

Reported Windows and macOS branches deliver a stealer or CloudSyncD.

What to look for

A script host or shell runs encoded content, a remote URL, or inline script syntax, then downloads or launches a payload.

Visibility: Defender-side logs where collected

Required logs

  • Windows process creation: Sysmon Event 1 or Defender DeviceProcessEvents
  • PowerShell script-block logging when enabled
  • Download / network and installer events

Coverage limit. Encoded commands and remote administration can be legitimate. This does not cover every script host, in-process execution, or non-Windows branch.

Windows process candidates. Sigma uses normalized process_creation fields; KQL requires Defender for Endpoint; SPL requires the Sysmon XML sourcetype. Tune collectors, fields and baselines before use.

Sigma
Draft SigmaDownload Sigma
title: Hunt - Suspicious Windows script execution
id: 2bb88802-9f03-4d99-a192-fe44c9484e8b
status: experimental
description: A script host or shell runs encoded content, a remote URL, or
  inline script syntax, then downloads or launches a payload.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
  - Authorized activity matching this broad hunt; baseline and investigate
    before alerting.
level: low
logsource:
  category: process_creation
  product: windows
detection:
  selection:
    Image|endswith:
      - \powershell.exe
      - \pwsh.exe
      - \mshta.exe
      - \wscript.exe
      - \cscript.exe
    CommandLine|contains:
      - -enc
      - http
      - FromBase64String
      - "javascript:"
      - "vbscript:"
  condition: selection
KQL
Draft KQLDownload KQL
// Draft hunt; requires Defender for Endpoint process telemetry for this OS.
DeviceProcessEvents
| where Timestamp > ago(7d)
| where FileName in~ ("powershell.exe", "pwsh.exe", "mshta.exe", "wscript.exe", "cscript.exe")
| where ProcessCommandLine contains "-enc" or ProcessCommandLine contains "http" or ProcessCommandLine contains "FromBase64String" or ProcessCommandLine contains "javascript:" or ProcessCommandLine contains "vbscript:"
| project Timestamp, DeviceName, AccountName, FileName, ProcessCommandLine, InitiatingProcessFileName, InitiatingProcessCommandLine, SHA1
| order by Timestamp desc
SPL
Draft SPLDownload SPL
sourcetype="XmlWinEventLog:Microsoft-Windows-Sysmon/Operational" EventCode=1
Image IN ("*\\powershell.exe", "*\\pwsh.exe", "*\\mshta.exe", "*\\wscript.exe", "*\\cscript.exe")
CommandLine IN ("*-enc*", "*http*", "*FromBase64String*", "*javascript:*", "*vbscript:*")
| table _time host User Image CommandLine ParentImage ParentCommandLine
| sort - _time
Investigation steps
  1. Follow the parent and child process tree and recover the script or URL from logs without executing it.
  2. Correlate the payload's signer/hash, persistence, and network activity; baseline approved automation.
Technical detailsTooling, impact & MITRE ATLAS

Attack context

AI's role
AI tool as lure
Additional role
AI infrastructure as C2
Initial access
Social engineering via fake accelerator / free AI credits; npm install of @apexacc/cli or @apexacc/audit

Tooling as reported

AI products, models and artifacts

  • Oh My Pi (repackaged as Apex Copilot)
  • Attacker LLM proxy arena.apexaccs.org/api/llm/v1
  • Attacker MCP arena.apexaccs.org/api/copilot/mcp
  • Fabricated 18-tool Apex Audit MCP
  • Targets Claude Desktop, Cursor, Windsurf configs

Other tooling

  • wscript.exe
  • PowerShell
  • UACME ComputerDefaults bypass
  • Go AMSI/ETW patcher
  • CloudSyncD (macOS)
  • Telegram exfil
  • bludness.exe / exegis.exe

Targets and reported impact

Targets / sectors
Cryptocurrency / Web3 founders and startups
Target geography
Multiple regions

Global Web3 targeting is described. A claimed presence at TOKEN2049 in Singapore is not evidence of a victim in Singapore.

Reported attribution

Elastic REF8801; Russian comments in history but nationality not established

Structured attribution

Actor
Elastic REF8801
Country
Russia

Detection rules

  • Review attack-specific indicatorsGUIDANCE

    Monitor for the techniques described in Apex Accelerator / REF8801. Refer to sources for specific IOCs and TTPs.

Source defensive notes

Block apexaccs.org / apexfdn.xyz / payload domains; audit MCP configs and OPENROUTER_BASE_URL; rotate AI keys; prevent wscript (Windows stealer chain depends on it)

MITRE ATLAS

AI-specific catalog annotations based on the reported behavior; these are not MITRE case-study assignments.

AML.T0081Modify AI Agent Configuration

The packages rewrite model-proxy and MCP settings. Source

AML.T0104Publish Poisoned AI Agent Tool

The accelerator distributes a rogue audit MCP tool. Source

Sources & research credit

1 source

Research by Jia Yu Chan, Daniel Stepanic · Elastic Security Labs

  1. elastic.co
Record & downloads
Source record name
Apex Accelerator / REF8801
Entry ID
apex-accelerator-ref8801-2026-10
Catalog review
October 20, 2026
Category
adversary-use-of-ai
AI involvement confidence
confirmed