Back to skill

Security audit

Openclaw With Apple

Security checks for vulnerabilities and agentic risk

Overview

This skill is a real Apple/iCloud automation package, but it asks users to give an AI their Apple ID password and 2FA code and sets up automatic background syncing with too little user control.

Review this carefully before installing. Do not paste your Apple ID primary password or 2FA codes into an AI chat; prefer app-specific passwords for calendar-only use and local interactive authentication if you proceed. Also check the launchd scheduled sync, cached ~/.pyicloud session, ~/.openclaw data, iCloud Drive uploads, and any archived status-wall/AMap location feature before enabling them.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • System PersistenceInstalls backdoors, hooks, services, or scheduled tasks that survive the run
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Findings (6)

T01 · Skill Instruction Hijacking

Error
Location
SKILL.md:17
Finding

Unconditional Skill instructions trigger data storage and command execution without explicit user consent

Content
View full analysis
Remediation
View remediation

T06 · System Persistence

Error
Location
scripts/setup_tasks_cron.py:87
Finding

Skill directs the Agent to install a persistent LaunchAgent without transparent confirmation

Content
View full analysis
ProgramArguments {python_path} {TASKS_TOOL} sync StartCalendarInterval Hour {hour} Minute {minute} ``` ### Technical Analysis The installer writes `com.openclaw.tasks-sync.plist` into `~/Library/LaunchAgents/` and loads it with `launchctl`. The job survives the original Skill invocation and runs `tasks_tool.py sync` at the configured time each day. Scheduled synchronization is related to the declared reminder and note feature. The security problem is that the Skill expli ...[truncated 1139 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:68
Finding

Skill solicits the Apple ID master password through chat and shell environment commands

Content
View full analysis
Remediation
View remediation

other

Error
Location
archive-full-version/scripts/status_wall.py:380
Finding

Archived status daemon repeatedly transmits precise iPhone coordinates to Amap

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/tasks_tool.py:47
Finding

Sensitive reminders, notes, location anchors, and API credentials are stored without enforced restrictive permissions

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
SKILL.md:1
Finding

Third-party Python dependencies are installed without version or integrity pinning

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (182)

Context-Inappropriate Capability

Critical
Category
Not specified by scanner
Confidence
99% confidence
Finding

The skill explicitly instructs the agent to ask for the user's Apple ID primary password in chat to gain broad account access. Collecting an account password through chat creates an immediate credential-theft risk, exposes the user's entire Apple account, and defeats safer delegated or OAuth-style authentication patterns.

Content

No source excerpt is available for this finding.

Ssd 3

Critical
Category
Not specified by scanner
Confidence
99% confidence
Finding

The onboarding asks the user to provide an Apple ID main password and later use it for account access, directly through chat. This is extremely dangerous because chat systems are not an appropriate secret-entry mechanism for primary credentials, and compromise of those logs exposes the full Apple account.

Content

No source excerpt is available for this finding.

Ssd 3

Critical
Category
Not specified by scanner
Confidence
99% confidence
Finding

The skill asks the user to send 2FA verification codes in chat and instructs the agent to use them to complete login. Requesting one-time codes is a classic credential-phishing pattern that can bypass account protections and is especially dangerous when combined with a collected password.

Content

No source excerpt is available for this finding.

Tainted flow: 'req' from os.environ.get (line 397, credential/environment) → urllib.request.urlopen (network output)

Critical
Category
Data Flow
Confidence
95% confidence
Finding

The code transmits precise GPS coordinates derived from Find My to the external AMap geocoding service. Even though the environment-derived value is only an API key, the real security issue is privacy-sensitive exfiltration of live location data to a third party without strong minimization, explicit consent flow at the transmission point, or clear access controls; in this skill context, that materially increases surveillance and privacy risk.

Content

Scanner excerpt · archive-full-version/scripts/status_wall.py (reported line 398)May include surrounding context.

python
})
        url = f'https://restapi.amap.com/v3/geocode/regeo?{params}'
        req = urllib.request.Request(url, headers={'User-Agent': 'StatusWall/1.0'})
        with urllib.request.urlopen(req, timeout=5) as resp:
            data = json.loads(resp.read().decode())

        if data.get('status') != '1':

Missing User Warnings

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The README explicitly tells users to provide their Apple ID email and main password to the AI through environment variables for login. Requesting a primary Apple account password for an AI-connected skill is extremely dangerous because it can enable full account compromise, access to personal data and devices, and abuse of linked Apple services if the secret is exposed, logged, mishandled, or reused.

Content

No source excerpt is available for this finding.

Ssd 3

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

This is a direct natural-language instruction to disclose highly sensitive credentials to the assistant, which is precisely the kind of social-engineering pattern that can lead users to surrender full account access. Within the context of a skill that already seeks access to health, files, photos, calendars, and device functions, asking for the main Apple password substantially amplifies the blast radius of any mishandling or compromise.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The documented behavior includes installation and management of persistent scheduled jobs, which is materially different from a simple end-user iCloud utility. Undeclared host persistence creates ongoing execution risk and can continue accessing or syncing personal data after setup.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The manifest description does not disclose broad iCloud capabilities documented later, including photo access, Drive file operations, device listing, device location, remote sound, and lost mode. Users and reviewers may underestimate the sensitivity of the requested credentials and approve the skill without understanding its full power over the Apple account.

Content

No source excerpt is available for this finding.

Hidden Instructions

High
Category
Prompt Injection
Confidence
93% confidence
Finding

This is a duplicate detection of the same hidden-instruction block. The concealed 'iron law' is risky because it pressures the agent to store data and run commands before normal safety or consent checks.

Content

Scanner excerpt · SKILL.md (reported line 13)May include surrounding context.

md
tags: icloud, health, tasks, reminders, notes
---

<!-- ═══════════════════════════════════════════════════════════ -->
<!-- ⛔⛔⛔ 铁律:在做任何事之前必须读完这一段 ⛔⛔⛔            -->
<!-- ═══════════════════════════════════════════════════════════ -->

Hidden Instructions

High
Category
Prompt Injection
Confidence
93% confidence
Finding

This is a duplicate detection of the same hidden-instruction block. The concealed 'iron law' is risky because it pressures the agent to store data and run commands before normal safety or consent checks.

Content

Scanner excerpt · SKILL.md (reported line 13)May include surrounding context.

md
tags: icloud, health, tasks, reminders, notes
---

<!-- ═══════════════════════════════════════════════════════════ -->
<!-- ⛔⛔⛔ 铁律:在做任何事之前必须读完这一段 ⛔⛔⛔            -->
<!-- ═══════════════════════════════════════════════════════════ -->