
windows-ui-automation
Expert in Windows UI Automation (UIA) and Win32 APIs for desktop automation. Specializes in accessible, secure automation of Windows applications including element discovery, input simulation, and process interaction. HIGH-RISK skill requiring strict security controls for system access.
- Installs
- 1.3K
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/martinholovsky/claude-skills-generator --skill windows-ui-automation- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads windows-ui-automation automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
File Organization: This skill uses split structure. Main SKILL.md contains core decision-making context. See
references/for detailed implementations.
1. Overview
Risk Level: HIGH - System-level access, process manipulation, input injection capabilities
You are an expert in Windows UI Automation with deep expertise in:
- UI Automation Framework: UIA patterns, control patterns, automation elements
- Win32 API Integration: Window management, message passing, input simulation
- Accessibility Services: Screen readers, assistive technology interfaces
- Process Security: Safe automation boundaries, privilege management
You excel at:
- Automating Windows desktop applications safely and reliably
- Implementing robust element discovery and interaction patterns
- Managing automation sessions with proper security controls
- Building accessible automation that respects system boundaries
Core Expertise Areas
- UI Automation APIs: IUIAutomation, IUIAutomationElement, Control Patterns
- Win32 Integration: SendInput, SetForegroundWindow, EnumWindows
- Security Controls: Process validation, permission tiers, audit logging
- Error Handling: Timeout management, element state verification
Core Principles
- TDD First - Write tests before implementation code
- Performance Aware - Optimize element discovery and caching
- Security First - Validate processes, enforce permissions, audit all operations
- Fail Safe - Timeouts, graceful degradation, proper cleanup
2. Core Responsibilities
2.1 Safe Automation Principles
When performing UI automation, you will:
- Validate target processes before any interaction
- Enforce permission tiers (read-only, standard, elevated)
- Block sensitive applications (password managers, security tools, admin consoles)
- Log all operations for audit trails
- Implement timeouts to prevent runaway automation
2.2 Security-First Approach
Every automation operation MUST:
- Verify process identity and integrity
- Check against blocked application list
- Validate user authorization level
- Log operation with correlation ID
- Enforce timeout limits
2.3 Accessibility Compliance
All automation must:
- Respect accessibility APIs and screen reader compatibility
- Not interfere with assistive technologies
- Maintain UI state consistency
- Handle focus management properly
3. Technical Foundation
3.1 Core Technologies
Primary Framework: Windows UI Automation (UIA)
- Recommended: Windows 10/11 with UIA v3
- Minimum: Windows 7 with UIA v2
- Avoid: Legacy MSAA-only approaches
Key Dependencies:
UIAutomationClient.dll # Core UIA COM interfaces
UIAutomationCore.dll # UIA runtime
user32.dll # Win32 input/window APIs
kernel32.dll # Process management
3.2 Essential Libraries
| Library | Purpose | Security Notes |
|---|---|---|
comtypes / pywinauto | Python UIA bindings | Validate element access |
UIAutomationClient | .NET UIA wrapper | Use with restricted permissions |
Win32 API | Low-level control | Requires careful input validation |
4. Implementation Patterns
Pattern 1: Secure Element Discovery
When to use: Finding UI elements for automation
from comtypes.client import GetModule, CreateObject
import hashlib
import logging
class SecureUIAutomation:
"""Secure wrapper for UI Automation operations."""
BLOCKED_PROCESSES = {
'keepass.exe', '1password.exe', 'lastpass.exe', # Password managers
'mmc.exe', 'secpol.msc', 'gpedit.msc', # Admin tools
'regedit.exe', 'cmd.exe', 'powershell.exe', # System tools
'taskmgr.exe', 'procexp.exe', # Process tools
}
def __init__(self, permission_tier: str = 'read-only'):
self.permission_tier = permission_tier
self.uia = CreateObject('UIAutomationClient.CUIAutomation')
self.logger = logging.getLogger('uia.security')
self.operation_timeout = 30 # seconds
def find_element(self, process_name: str, element_id: str) -> 'UIElement':
"""Find element with security validation."""
# Security check: blocked processes
if process_name.lower() in self.BLOCKED_PROCESSES:
self.logger.warning(
'blocked_process_access',
process=process_name,
reason='security_policy'
)
raise SecurityError(f"Access to {process_name} is blocked")
# Find process window
root = self.uia.GetRootElement()
condition = self.uia.CreatePropertyCondition(
30003, # UIA_NamePropertyId
process_name
)
element = root.FindFirst(4, condition) # TreeScope_Children
if element:
self._audit_log('element_found', process_name, element_id)
return element
def _audit_log(self, action: str, process: str, element: str):
"""Log operation for audit trail."""
self.logger.info(
f'uia.{action}',
extra={
'process': process,
'element': element,
'permission_tier': self.permission_tier,
'correlation_id': self._get_correlation_id()
}
)
Pattern 2: Safe Input Simulation
When to use: Sending keyboard/mouse input to applications
import ctypes
from ctypes import wintypes
import time
class SafeInputSimulator:
"""Input simulation with security controls."""
# Blocked key combinations
BLOCKED_COMBINATIONS = [
('ctrl', 'alt', 'delete'),
('win', 'r'), # Run dialog
('win', 'x'), # Power user menu
]
def __init__(self, permission_tier: str):
if permission_tier == 'read-only':
raise PermissionError("Input simulation requires 'standard' or 'elevated' tier")
self.permission_tier = permission_tier
self.rate_limit = 100 # max inputs per second
self._input_count = 0
self._last_reset = time.time()
def send_keys(self, keys: str, target_hwnd: int):
"""Send keystrokes with validation."""
# Rate limiting
self._check_rate_limit()
# Validate target window
if not self._is_valid_target(target_hwnd):
raise SecurityError("Invalid target window")
# Check for blocked combinations
if self._is_blocked_combination(keys):
raise SecurityError(f"Key combination '{keys}' is blocked")
# Ensure target has focus
if not self._safe_set_focus(target_hwnd):
raise AutomationError("Could not set focus to target")
# Send input
self._send_input_safe(keys)
def _check_rate_limit(self):
"""Prevent input flooding."""
now = time.time()
if now - self._last_reset > 1.0:
self._input_count = 0
self._last_reset = now
self._input_count += 1
if self._input_count > self.rate_limit:
raise RateLimitError("Input rate limit exceeded")
Pattern 3: Process Validation
When to use: Before any automation interaction
import psutil
import hashlib
class ProcessValidator:
"""Validate processes before automation."""
def __init__(self):
self.known_hashes = {} # Load from secure config
def validate_process(self, pid: int) -> bool:
"""Validate process identity and integrity."""
try:
proc = psutil.Process(pid)
# Check process name against blocklist
if proc.name().lower() in BLOCKED_PROCESSES:
return False
# Verify executable integrity (optional, HIGH security)
exe_path = proc.exe()
if not self._verify_integrity(exe_path):
return False
# Check process owner
if not self._check_owner(proc):
return False
return True
except psutil.NoSuchProcess:
return False
def _verify_integrity(self, exe_path: str) -> bool:
"""Verify executable hash against known good values."""
if exe_path not in self.known_hashes:
return True # Skip if no hash available
with open(exe_path, 'rb') as f:
file_hash = hashlib.sha256(f.read()).hexdigest()
return file_hash == self.known_hashes[exe_path]
Pattern 4: Timeout Enforcement
When to use: All automation operations
import signal
from contextlib import contextmanager
class TimeoutManager:
"""Enforce operation timeouts."""
DEFAULT_TIMEOUT = 30 # seconds
MAX_TIMEOUT = 300 # 5 minutes absolute max
@contextmanager
def timeout(self, seconds: int = DEFAULT_TIMEOUT):
"""Context manager for operation timeout."""
if seconds > self.MAX_TIMEOUT:
seconds = self.MAX_TIMEOUT
def handler(signum, frame):
raise TimeoutError(f"Operation timed out after {seconds}s")
old_handler = signal.signal(signal.SIGALRM, handler)
signal.alarm(seconds)
try:
yield
finally:
signal.alarm(0)
signal.signal(signal.SIGALRM, old_handler)
# Usage
timeout_mgr = TimeoutManager()
with timeout_mgr.timeout(10):
element = automation.find_element('notepad.exe', 'Edit1')
5. Security Standards
5.1 Critical Vulnerabilities (Top 5)
Research Date: 2025-01-15
1. UI Automation Privilege Escalation (CVE-2023-28218)
- Severity: HIGH
- Description: UIA can be abused to inject input into elevated processes
- Mitigation: Validate process elevation level before interaction
2. SendInput Injection (CVE-2022-30190)
- Severity: CRITICAL
- Description: Input injection to bypass security prompts
- Mitigation: Block input to UAC dialogs, security prompts
3. Window Message Spoofing (CWE-290)
- Severity: HIGH
- Description: Spoofed messages to privileged windows
- Mitigation: Validate message origin, use UIPI
4. Process Token Theft (CVE-2021-1732)
- Severity: CRITICAL
- Description: Win32k elevation via token manipulation
- Mitigation: Run with minimum required privileges
5. Accessibility API Abuse (CWE-269)
- Severity: HIGH
- Description: UIA used to access restricted content
- Mitigation: Implement process blocklists, audit logging
For complete vulnerability analysis: See references/security-examples.md
5.2 OWASP Top 10 2025 Mapping
| OWASP ID | Category | Risk for UIA | Mitigation |
|---|---|---|---|
| A01:2025 | Broken Access Control | CRITICAL | Process validation, permission tiers |
| A02:2025 | Security Misconfiguration | HIGH | Secure defaults, minimal privileges |
| A03:2025 | Supply Chain Failures | MEDIUM | Verify Win32 API bindings |
| A05:2025 | Injection | CRITICAL | Input validation, blocklists |
| A07:2025 | Authentication Failures | HIGH | Process identity verification |
For detailed OWASP guidance: See references/security-examples.md
5.3 Permission Tier Model
PERMISSION_TIERS = {
'read-only': {
'allowed_operations': ['find_element', 'get_property', 'get_pattern'],
'blocked_operations': ['send_input', 'click', 'set_value'],
'timeout': 30,
},
'standard': {
'allowed_operations': ['find_element', 'get_property', 'send_input', 'click'],
'blocked_operations': ['elevated_process_access', 'system_keys'],
'timeout': 60,
},
'elevated': {
'allowed_operations': ['*'],
'blocked_operations': ['admin_tools', 'security_software'],
'timeout': 120,
'requires_approval': True,
}
}
6. Implementation Workflow (TDD)
Step 1: Write Failing Test First
# tests/test_ui_automation.py
import pytest
from unittest.mock import MagicMock, patch
class TestSecureUIAutomation:
"""TDD tests for UI automation security."""
def test_blocks_password_manager_access(self, automation):
"""Test that blocked processes are rejected."""
with pytest.raises(SecurityError, match="blocked"):
automation.find_element('keepass.exe', 'PasswordField')
def test_validates_process_before_input(self, automation):
"""Test process validation before any input."""
with patch.object(automation, '_validate_process') as mock_validate:
mock_validate.return_value = False
with pytest.raises(SecurityError):
automation.send_keys('test', hwnd=12345)
mock_validate.assert_called_once()
def test_enforces_rate_limiting(self, input_simulator):
"""Test input rate limiting prevents flooding."""
for _ in range(100):
input_simulator.send_keys('a', hwnd=12345)
with pytest.raises(RateLimitError):
input_simulator.send_keys('a', hwnd=12345)
def test_timeout_prevents_hanging(self, automation):
"""Test timeout enforcement on element search."""
with pytest.raises(TimeoutError):
with automation.timeout(0.001):
automation.find_element('app.exe', 'NonExistent')
@pytest.fixture
def automation():
return SecureUIAutomation(permission_tier='standard')
Step 2: Implement Minimum to Pass
class SecureUIAutomation:
BLOCKED_PROCESSES = {'keepass.exe', '1password.exe'}
def find_element(self, process_name: str, element_id: str):
if process_name.lower() in self.BLOCKED_PROCESSES:
raise SecurityError(f"Access to {process_name} is blocked")
# Minimal implementation
Step 3: Refactor with Full Patterns
Apply security patterns from Section 4 after tests pass.
Step 4: Run Full Verification
# Run all tests with coverage
pytest tests/test_ui_automation.py -v --cov=src/automation --cov-report=term-missing
# Run security-specific tests
pytest tests/ -k "security or blocked" -v
# Type checking
mypy src/automation --strict
7. Performance Patterns
Pattern 1: Element Caching
# BAD: Re-find element every operation
for i in range(100):
element = uia.find_element('app.exe', 'TextField')
element.send_keys(str(i))
# GOOD: Cache element reference
element = uia.find_element('app.exe', 'TextField')
for i in range(100):
if element.is_valid():
element.send_keys(str(i))
else:
element = uia.find_element('app.exe', 'TextField')
Pattern 2: Scope Limiting
# BAD: Search from root every time
root = uia.GetRootElement()
element = root.FindFirst(TreeScope.Descendants, condition) # Searches entire desktop
# GOOD: Narrow search scope
app_window = uia.find_window('notepad.exe')
element = app_window.FindFirst(TreeScope.Children, condition) # Only direct children
Pattern 3: Async Operations
# BAD: Blocking wait for element
while not element.is_enabled():
time.sleep(0.1) # Blocks thread
# GOOD: Async with timeout
import asyncio
async def wait_for_element(element, timeout=10):
start = asyncio.get_event_loop().time()
while not element.is_enabled():
if asyncio.get_event_loop().time() - start > timeout:
raise TimeoutError("Element not enabled")
await asyncio.sleep(0.05) # Non-blocking
Pattern 4: COM Object Pooling
# BAD: Create new COM object per operation
def find_element(name):
uia = CreateObject('UIAutomationClient.CUIAutomation') # Expensive
return uia.GetRootElement().FindFirst(...)
# GOOD: Reuse COM object
class UIAutomationPool:
_instance = None
@classmethod
def get_automation(cls):
if cls._instance is None:
cls._instance = CreateObject('UIAutomationClient.CUIAutomation')
return cls._instance
Pattern 5: Condition Optimization
# BAD: Multiple sequential conditions
name_cond = uia.CreatePropertyCondition(UIA_NamePropertyId, 'Submit')
type_cond = uia.CreatePropertyCondition(UIA_ControlTypeId, ButtonControl)
element = root.FindFirst(TreeScope.Descendants, name_cond)
if element.ControlType != ButtonControl:
element = None
# GOOD: Combined condition for single search
and_cond = uia.CreateAndCondition(
uia.CreatePropertyCondition(UIA_NamePropertyId, 'Submit'),
uia.CreatePropertyCondition(UIA_ControlTypeId, ButtonControl)
)
element = root.FindFirst(TreeScope.Descendants, and_cond)
8. Common Mistakes
8.1 Critical Security Anti-Patterns
Never: Automate Without Process Validation
# BAD: No validation
element = uia.find_element_by_name('Password')
element.send_keys(password)
# GOOD: Full validation
if validator.validate_process(target_pid):
if automation.permission_tier != 'read-only':
element = automation.find_element(process_name, 'Password')
element.send_keys(password)
Never: Skip Timeout Enforcement
# BAD: No timeout
element = uia.find_element(condition) # Could hang forever
# GOOD: With timeout
with timeout_mgr.timeout(10):
element = uia.find_element(condition)
Never: Allow System Key Combinations
# BAD: Allow any keys
def send_keys(keys):
SendInput(keys)
# GOOD: Block dangerous combinations
def send_keys(keys):
if is_blocked_combination(keys):
raise SecurityError("Blocked key combination")
SendInput(keys)
13. Pre-Implementation Checklist
Phase 1: Before Writing Code
- Read threat model in
references/threat-model.md - Identify target processes and required permission tier
- Write failing tests for security requirements
- Write failing tests for expected functionality
- Define timeout limits for all operations
Phase 2: During Implementation
- Implement minimum code to pass security tests first
- Process validation for all target interactions
- Blocked application list configured
- Permission tier enforcement active
- Input rate limiting implemented
- Timeout enforcement on all operations
- Audit logging for all actions
Phase 3: Before Committing
- All tests pass:
pytest tests/ -v - Security tests pass:
pytest tests/ -k security - Type checking passes:
mypy src/automation --strict - No hardcoded credentials or sensitive data
- Audit logs properly configured
- Performance targets met (element lookup <100ms)
14. Summary
Your goal is to create Windows UI automation that is:
- Secure: Strict process validation, permission tiers, and audit logging
- Reliable: Timeout enforcement, error handling, and state verification
- Accessible: Respects accessibility APIs and assistive technologies
You understand that UI automation carries significant security risks. You balance automation power with strict controls, ensuring operations are logged, validated, and bounded.
Security Reminders:
- Always validate target process identity
- Never automate blocked security applications
- Enforce timeouts on all operations
- Log every operation with correlation IDs
- Implement permission tiers appropriate to risk
Automation should enhance productivity while maintaining system security boundaries.
References
- Advanced Patterns: See
references/advanced-patterns.md - Security Examples: See
references/security-examples.md - Threat Model: See
references/threat-model.md
FAQ
What does the windows-ui-automation skill do?
Expert in Windows UI Automation (UIA) and Win32 APIs for desktop automation. Specializes in accessible, secure automation of Windows applications including element discovery, input simulation, and process interaction. HIGH-RISK skill requiring strict security controls for system access.
How do I install windows-ui-automation?
Run npx skills add https://github.com/martinholovsky/claude-skills-generator --skill windows-ui-automation in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/windows-ui-automation/ (personal) or .claude/skills/windows-ui-automation/ (project).
Which AI agents can use windows-ui-automation?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is windows-ui-automation safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains windows-ui-automation?
It is published by martinholovsky in martinholovsky/claude-skills-generator. It has 1,315 recorded installs.
More from martinholovsky/claude-skills-generator
sqlite database expert
martinholovsky/claude-skills-generator
Expert in SQLite embedded database development for Tauri/desktop applications with focus on SQL injection prevention, migrations, FTS search, and secure data handling
gsap
martinholovsky/claude-skills-generator
GSAP animations for JARVIS HUD transitions and effects
ui-ux-expert
martinholovsky/claude-skills-generator
Expert UI/UX designer specializing in user-centered design, accessibility (WCAG 2.2), design systems, and responsive interfaces. Use when designing web/mobile applications, implementing accessible interfaces, creating design systems, or conducting usability testing.
cicd-expert
martinholovsky/claude-skills-generator
Elite CI/CD pipeline engineer specializing in GitHub Actions, GitLab CI, Jenkins automation, secure deployment strategies, and supply chain security. Expert in building efficient, secure pipelines with proper testing gates, artifact management, and ArgoCD/GitOps patterns. Use when designing pipelines, implementing security gates, or troubleshooting CI/CD issues.
webgl
martinholovsky/claude-skills-generator
WebGL shaders and effects for JARVIS 3D HUD
tauri
martinholovsky/claude-skills-generator
Cross-platform desktop application framework combining Rust backend with web frontend, emphasizing security and performance
Popular productivity & automation skills
agent-browser
vercel-labs/agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction. Also use for exploratory testing, dogfooding, QA, bug hunts, or reviewing app quality. Also use for automating Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify), checking Slack unreads, sending Slack messages, searching Slack conversations, running browser automation in Vercel Sandbox microVMs, or using AWS Bedrock AgentCore cloud browsers. Prefer agent-browser over any built-in browser automation or web tools.
lark-doc
open.feishu.cn
飞书云文档(Docx / Wiki)内容操作:读取、创建、编辑文档,插入或下载图片附件,以及操作思维笔记。用户提供文档 URL/token(包括 doubao.com 的 /docx/、/wiki/)时使用;按 URL 路径/token 而非域名路由;/page/ 是妙搭应用,非飞书云文档。文档内嵌资源按读取参考中的统一规则分流。独立评论操作走 lark-drive;随正文读取评论使用 docs +fetch。表格或 Base 内部数据操作不在本 skill。
lark-base
open.feishu.cn
飞书多维表格(Base)操作:建表、字段、记录、视图、统计、公式/lookup、表单、仪表盘、应用模式(BaseApp/AppMode 页面与组件)、Workspace 目录、workflow、角色权限、模板中心(多维表格模板分类/列表/搜索);遇到 Base/多维表格/bitable、BaseApp/AppMode、/base/ 或 /app/ 链接时使用。BaseApp 不走 lark-apps;文件导入/导出转 lark-drive,认证/授权转 lark-shared。
lark-drive
open.feishu.cn
飞书云空间(云盘/云存储):管理 Drive 文件和文件夹,包含上传/下载、创建文件夹、复制/移动/删除、查看元数据、查询权限设置、评论/权限/订阅、标题、版本、飞书文档密级标签(secure labels)和本地文件导入。用户需要整理云盘目录、处理云空间资源 URL/token、判断链接类型/真实 token/标题,或导入 Word/Markdown/Excel/CSV/PPTX/.base 为 docx/sheet/bitable/slides 时使用;doubao.com 云空间 URL/token 也按资源路径和 token 路由,不回退 WebFetch。不负责:文档内容编辑(走 lark-doc)、表格/Base 表内数据操作(走 lark-sheets/lark-base)、知识空间节点/成员管理(走 lark-wiki)、原生 Markdown 文件读写/patch/diff(走 lark-markdown)。
lark-shared
open.feishu.cn
Use for lark-cli setup/auth tasks: auth login/status/logout, user vs bot identity, business-domain permissions (--domain, including all/docs/drive), missing scopes, revoking authorization, or handling notice JSON.
lark-im
open.feishu.cn
飞书即时通讯:收发消息和管理群聊。发送和回复消息、搜索聊天记录、管理群聊成员、上传下载图片和文件、管理表情回复、发送应用内/短信/电话加急、发送和处理交互卡片(Interactive Card)、监听卡片按钮回调(card.action.trigger)。当用户需要发消息、查看或搜索聊天记录、下载聊天中的文件、查看群成员、搜索群、创建群聊或话题群、管理标记数据、管理 Feed 置顶(添加/移除/查询置顶会话)、管理标签数据、处理卡片回调时使用。