GhostAction Stage 2 stealer targeting cloud and AI credentials
Malicious workflows and a VS Code auto-run task steal AWS, npm, SSH and Claude/Gemini credentials across 94 repos. AI is a credential target here, not the attack operator.
What happened
GhostAction pushes malicious GitHub Actions then Stage 2: .vscode/tasks.json folderOpen runs node on fake fa-solid-400.woff2 JS stealer; cron */5 pulls C2. Targets AWS/npm/git/kube/SSH plus Claude/Gemini creds. 94 repos / 66 owners. PolinRider copycat technique, not DPRK per OSM.
Steals AI client credentials (~/.claude/.credentials.json, ~/.gemini/oauth_creds.json, sk-ant- grep) alongside cloud/git tokens
How it happened
Read the flow as text & sources
A catalog reconstruction of the supplied source account. Primary attack is CI secret theft + VS Code folderOpen stealer; AI is a credential target (Claude/Gemini keys), not the attack operator. Included because Jose's post calls out AI tooling theft.
- 01 · Compromised account changes a repository
Malicious GitHub Actions and workspace tasks are introduced.
Hunt: A package or workflow change introduces a lifecycle hook, remote dependency or new command that accesses secrets or external services.
Package installation and unfamiliar names are not malicious by themselves. Dependency resolution and CI audit records are required.
Source for step 1 - 02 · Trusted workspace opens a task
VS Code executes Node content disguised as a font file.
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 - 03 · AI credential stores are searched
The stealer targets Claude and Gemini tokens alongside cloud and developer credentials.
Hunt: An unexpected process or tool accesses credential stores and is followed by token use or outbound transfer inconsistent with the owner.
Defender DeviceFileEvents is not a general file-read audit trail. Collect suitable object-access, EDR or application logs; legitimate developer tools also read secrets.
Source for step 3 - 04 · Secrets are stolen with persistence
The report describes periodic C2 retrieval through cron and a multi-repository campaign.
Hunt: A task, Run key, cron entry, shell profile, or startup file is changed and later launches the same unexpected program.
Windows and Unix persistence require different collectors. A configuration change without a subsequent launch is not proof that persistence ran.
Source for step 4
Why it matters
Reported entry point: Compromised GitHub account → malicious workflow + VS Code trusted workspace auto task. Reported impact: The report describes periodic C2 retrieval through cron and a multi-repository campaign; reported targets include GitHub maintainers / developers (many dormant repos).
Check these source-linked conditions against your setup.
The reported stealer targets local development, cloud and Claude / Gemini credentials.
Conditions to check- The malicious workflow or VS Code folder-open task must run with access to locally stored credentials.
- AI is a credential target in this account; the report does not establish an AI-operated intrusion.
Source-linked applicability guidance. Confirm current scope in the original report.
What you can do
Protection & prevention
Review workflow and folder-open task changes, then rotate exposed development and AI credentials.
task.allowAutomaticTasks off; Restricted Mode; crontab -l for #ghc2; rotate Claude/Gemini/cloud/git creds
Catalog defensive guidance based on the reported behavior. Confirm scope and current mitigations in the original source.
Detection queries
Draft query starters grouped by the attack stages where they apply.
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.
01Compromised account changes a repositoryDependency / CI trust boundaryCustom schema · Draft
Malicious GitHub Actions and workspace tasks are introduced.
What to look for
A package or workflow change introduces a lifecycle hook, remote dependency or new command that accesses secrets or external services.
Visibility: Defender-side logs where collected
Required logs
- Package-manager and CI job logs
- Lockfile / manifest / workflow diffs
- Runner process trees, secret access and egress
Coverage limit. Package installation and unfamiliar names are not malicious by themselves. Dependency resolution and CI audit records are required.
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
title: Hunt - Dependency / CI trust boundary
id: 5fa9c45a-ac1c-4666-a598-e80626320997
status: experimental
description: A package or workflow change introduces a lifecycle hook, remote
dependency or new command that accesses secrets or external services.
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:
- package.install
- ci.workflow.change
- package.lifecycle.exec
condition: selection
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 ("package.install", "ci.workflow.change", "package.lifecycle.exec")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated descSPL
sourcetype="openattacks:audit"
EventType IN ("package.install", "ci.workflow.change", "package.lifecycle.exec")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _timeInvestigation steps
- Compare package provenance, publish identity, lifecycle scripts and resolved dependency URLs with the approved version.
- Correlate the install or workflow to secret reads and subsequent token use; distinguish exposure from confirmed compromise.
02Trusted workspace opens a taskAgent trust / configuration changesCustom schema · Draft
VS Code executes Node content disguised as a font file.
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
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
// 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 descSPL
sourcetype="openattacks:audit"
EventType IN ("agent.config.write", "agent.approval.change")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _timeInvestigation steps
- Diff command, args, environment, origin, autoApprove and task settings against the previously approved version.
- Correlate the author and change with the next agent launch; inspect path canonicalization and trailing-dot variants where relevant.
03AI credential stores are searchedSensitive store access → credential useCustom schema · Draft
The stealer targets Claude and Gemini tokens alongside cloud and developer credentials.
What to look for
An unexpected process or tool accesses credential stores and is followed by token use or outbound transfer inconsistent with the owner.
Visibility: Defender-side logs where collected
Required logs
- File-read / API access telemetry for relevant secret stores
- Process identity and command lines
- Identity, token and outbound-network audit logs
Coverage limit. Defender DeviceFileEvents is not a general file-read audit trail. Collect suitable object-access, EDR or application logs; legitimate developer tools also read secrets.
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
title: Hunt - Sensitive store access → credential use
id: 46c46c62-2d83-4198-a8bf-d6637094527b
status: experimental
description: An unexpected process or tool accesses credential stores and is
followed by token use or outbound transfer inconsistent with the owner.
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:
- secret.read
- credential.store.read
condition: selection
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 ("secret.read", "credential.store.read")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated descSPL
sourcetype="openattacks:audit"
EventType IN ("secret.read", "credential.store.read")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _timeInvestigation steps
- Identify the actor, process and exact store touched, then pivot to token use from new devices or infrastructure.
- Correlate reads with egress and rotate affected credentials after confirming exposure.
04Secrets are stolen with persistenceNew persistence → recurring executionCustom schema · Draft
The report describes periodic C2 retrieval through cron and a multi-repository campaign.
What to look for
A task, Run key, cron entry, shell profile, or startup file is changed and later launches the same unexpected program.
Visibility: Defender-side logs where collected
Required logs
- Platform-specific registry, scheduled-task, cron and startup-file changes
- Process creation and file provenance
- User / administrative change records
Coverage limit. Windows and Unix persistence require different collectors. A configuration change without a subsequent launch is not proof that persistence ran.
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
title: Hunt - New persistence → recurring execution
id: 6154ff56-2978-45fa-aecd-86d91557f32e
status: experimental
description: A task, Run key, cron entry, shell profile, or startup file is
changed and later launches the same unexpected program.
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:
- persistence.create
- persistence.modify
condition: selection
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 ("persistence.create", "persistence.modify")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated descSPL
sourcetype="openattacks:audit"
EventType IN ("persistence.create", "persistence.modify")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _timeInvestigation steps
- Diff the persistence artifact and identify the creating process / principal.
- Correlate the subsequent launch and network behavior; baseline deployment and management agents.
Technical detailsTooling, impact & MITRE ATLAS
Attack context
- AI's role
- AI credential theft
- Initial access
- Compromised GitHub account → malicious workflow + VS Code trusted workspace auto task
Tooling as reported
AI products, models and artifacts
Other tooling
- GitHub Actions
- VS Code tasks.json
- node
- cron
- 193.32.204.199:8080
Targets and reported impact
- Targets / sectors
- GitHub maintainers / developers (many dormant repos)
- Reported scale / exposure
- 94 compromised repos; Stage 1 Oct 8, Stage 2 Oct 15 2026
Reported attribution
GhostAction; PolinRider-technique copycat (no shared IOCs with DPRK per OSM)
Structured attribution
- Actor
- GhostAction
Detection rules
- Review attack-specific indicatorsGUIDANCE
Monitor for the techniques described in GhostAction Stage 2 stealer (AI credential theft). Refer to sources for specific IOCs and TTPs.
Source defensive notes
task.allowAutomaticTasks off; Restricted Mode; crontab -l for #ghc2; rotate Claude/Gemini/cloud/git creds
MITRE ATLAS
AI-specific catalog annotations based on the reported behavior; these are not MITRE case-study assignments.
The stealer searches local AI credential stores. Source
Sources & research credit
1 sourceResearch by Thijs Xhaflaire (Jamf Threat Labs); OpenSourceMalware deep dive · Jamf Threat Labs / OpenSourceMalware
Record & downloads
- Source record name
- GhostAction Stage 2 stealer (AI credential theft)
- Entry ID
- ghostaction-ai-creds-2026-10
- Catalog review
- October 20, 2026
- Category
- attacks-on-ai
- AI involvement confidence
- unverified