claude-mpm 4.17.0__py3-none-any.whl → 4.18.3__py3-none-any.whl

claude_mpm/skills/bundled/pdf.md

@@ -0,0 +1,141 @@
+---
+skill_id: pdf
+skill_version: 0.1.0
+description: Common PDF operations and libraries across languages.
+updated_at: 2025-10-30T17:00:00Z
+tags: [pdf, document-processing, manipulation, media]
+---
+
+# PDF Manipulation
+
+Common PDF operations and libraries across languages.
+
+## Python (PyPDF2 / pikepdf)
+
+### Reading PDF
+```python
+from PyPDF2 import PdfReader
+
+reader = PdfReader("document.pdf")
+print(f"Pages: {len(reader.pages)}")
+
+# Extract text
+text = reader.pages[0].extract_text()
+```
+
+### Writing PDF
+```python
+from PyPDF2 import PdfWriter, PdfReader
+
+reader = PdfReader("input.pdf")
+writer = PdfWriter()
+
+# Copy pages
+for page in reader.pages:
+    writer.add_page(page)
+
+# Save
+with open("output.pdf", "wb") as f:
+    writer.write(f)
+```
+
+### Merging PDFs
+```python
+from PyPDF2 import PdfMerger
+
+merger = PdfMerger()
+merger.append("file1.pdf")
+merger.append("file2.pdf")
+merger.write("merged.pdf")
+merger.close()
+```
+
+### Splitting PDF
+```python
+reader = PdfReader("input.pdf")
+for i, page in enumerate(reader.pages):
+    writer = PdfWriter()
+    writer.add_page(page)
+    with open(f"page_{i}.pdf", "wb") as f:
+        writer.write(f)
+```
+
+## JavaScript (pdf-lib)
+
+```javascript
+import { PDFDocument } from 'pdf-lib';
+import fs from 'fs';
+
+// Load existing PDF
+const existingPdfBytes = fs.readFileSync('input.pdf');
+const pdfDoc = await PDFDocument.load(existingPdfBytes);
+
+// Get pages
+const pages = pdfDoc.getPages();
+const firstPage = pages[0];
+
+// Add text
+firstPage.drawText('Hello World!', {
+  x: 50,
+  y: 50,
+  size: 30
+});
+
+// Save
+const pdfBytes = await pdfDoc.save();
+fs.writeFileSync('output.pdf', pdfBytes);
+```
+
+## Common Operations
+
+### Extracting Images
+```python
+import fitz  # PyMuPDF
+
+doc = fitz.open("document.pdf")
+for page_num in range(len(doc)):
+    page = doc[page_num]
+    images = page.get_images()
+
+    for img_index, img in enumerate(images):
+        xref = img[0]
+        base_image = doc.extract_image(xref)
+        image_bytes = base_image["image"]
+
+        with open(f"image_{page_num}_{img_index}.png", "wb") as f:
+            f.write(image_bytes)
+```
+
+### Adding Watermark
+```python
+from PyPDF2 import PdfReader, PdfWriter
+
+# Create watermark PDF first
+watermark_reader = PdfReader("watermark.pdf")
+watermark_page = watermark_reader.pages[0]
+
+# Apply to document
+reader = PdfReader("input.pdf")
+writer = PdfWriter()
+
+for page in reader.pages:
+    page.merge_page(watermark_page)
+    writer.add_page(page)
+
+with open("watermarked.pdf", "wb") as f:
+    writer.write(f)
+```
+
+### Compressing PDF
+```python
+from pikepdf import Pdf
+
+with Pdf.open("input.pdf") as pdf:
+    pdf.save("compressed.pdf", compress_streams=True)
+```
+
+## Remember
+- Check PDF file size before processing
+- Handle corrupted PDFs gracefully
+- Use appropriate library for task (pikepdf > PyPDF2 for complex ops)
+- Consider memory usage for large PDFs

claude_mpm/skills/bundled/performance-profiling.md

@@ -0,0 +1,567 @@
+---
+skill_id: performance-profiling
+skill_version: 0.1.0
+description: Systematic approach to identifying and optimizing performance bottlenecks, eliminating redundant profiling guidance per agent.
+updated_at: 2025-10-30T17:00:00Z
+tags: [performance, profiling, optimization, benchmarking]
+---
+
+# Performance Profiling
+
+Systematic approach to identifying and optimizing performance bottlenecks. Eliminates ~200-250 lines of redundant profiling guidance per agent.
+
+## Core Principle: Measure Before Optimizing
+
+**Never optimize without profiling data.** Intuition about performance is usually wrong.
+
+> "Premature optimization is the root of all evil" - Donald Knuth
+
+## The Profiling Process
+
+### 1. Establish Baseline
+
+Before any optimization:
+```
+□ Measure current performance
+□ Define acceptable performance targets
+□ Identify critical code paths
+□ Document current metrics
+```
+
+### 2. Profile to Find Bottlenecks
+
+Use profiling tools to identify hot spots:
+```
+□ CPU profiling (where time is spent)
+□ Memory profiling (allocation patterns)
+□ I/O profiling (disk/network bottlenecks)
+□ Database query profiling
+```
+
+### 3. Optimize Targeted Areas
+
+Focus on the biggest bottlenecks first (80/20 rule).
+
+### 4. Measure Impact
+
+Verify that optimizations actually improved performance.
+
+## Language-Specific Profiling
+
+### Python
+
+#### CPU Profiling with cProfile
+
+```python
+import cProfile
+import pstats
+from pstats import SortKey
+
+# Profile a function
+profiler = cProfile.Profile()
+profiler.enable()
+
+slow_function()
+
+profiler.disable()
+
+# Analyze results
+stats = pstats.Stats(profiler)
+stats.sort_stats(SortKey.CUMULATIVE)
+stats.print_stats(20)  # Top 20 functions by cumulative time
+
+# Output shows:
+# - ncalls: number of calls
+# - tottime: total time in function (excluding subcalls)
+# - cumtime: cumulative time (including subcalls)
+```
+
+#### Line-by-Line Profiling
+
+```python
+from line_profiler import LineProfiler
+
+@profile  # Decorator for line_profiler
+def bottleneck_function():
+    data = load_data()
+    processed = process_data(data)
+    result = analyze(processed)
+    return result
+
+# Run: kernprof -l -v script.py
+# Shows time spent on each line
+```
+
+#### Memory Profiling
+
+```python
+from memory_profiler import profile
+
+@profile
+def memory_intensive():
+    large_list = [i for i in range(10000000)]  # Shows memory spike
+    processed = [x * 2 for x in large_list]    # Shows peak memory
+    return sum(processed)
+
+# Run: python -m memory_profiler script.py
+# Shows line-by-line memory usage
+```
+
+### JavaScript/Node.js
+
+#### Built-in Profiling
+
+```javascript
+// CPU profiling
+console.profile('MyOperation');
+expensiveOperation();
+console.profileEnd('MyOperation');
+
+// Time measurement
+console.time('operation');
+performOperation();
+console.timeEnd('operation');
+```
+
+#### Node.js Performance API
+
+```javascript
+const { performance, PerformanceObserver } = require('perf_hooks');
+
+// Mark specific points
+performance.mark('start-operation');
+performOperation();
+performance.mark('end-operation');
+
+// Measure duration
+performance.measure('operation-duration', 'start-operation', 'end-operation');
+
+const observer = new PerformanceObserver((items) => {
+  items.getEntries().forEach((entry) => {
+    console.log(`${entry.name}: ${entry.duration}ms`);
+  });
+});
+
+observer.observe({ entryTypes: ['measure'] });
+```
+
+#### Chrome DevTools Profiling
+
+```javascript
+// Run with: node --inspect script.js
+// Open chrome://inspect in Chrome
+// Use Performance tab to record and analyze
+
+function expensiveOperation() {
+  // CPU-intensive work
+  for (let i = 0; i < 1000000; i++) {
+    Math.sqrt(i);
+  }
+}
+
+expensiveOperation();
+```
+
+### Go
+
+#### CPU Profiling
+
+```go
+import (
+    "os"
+    "runtime/pprof"
+)
+
+func main() {
+    // Start CPU profiling
+    f, _ := os.Create("cpu.prof")
+    defer f.Close()
+    pprof.StartCPUProfile(f)
+    defer pprof.StopCPUProfile()
+
+    // Code to profile
+    expensiveOperation()
+
+    // Analyze with: go tool pprof cpu.prof
+}
+```
+
+#### Memory Profiling
+
+```go
+import (
+    "os"
+    "runtime/pprof"
+)
+
+func main() {
+    // Code to profile
+    expensiveOperation()
+
+    // Write memory profile
+    f, _ := os.Create("mem.prof")
+    defer f.Close()
+    pprof.WriteHeapProfile(f)
+
+    // Analyze with: go tool pprof mem.prof
+}
+```
+
+#### Benchmarking
+
+```go
+func BenchmarkExpensiveOperation(b *testing.B) {
+    for i := 0; i < b.N; i++ {
+        expensiveOperation()
+    }
+}
+
+// Run: go test -bench=. -benchmem
+// Shows:
+// - ns/op: nanoseconds per operation
+// - B/op: bytes allocated per operation
+// - allocs/op: allocations per operation
+```
+
+### Rust
+
+#### Benchmarking with Criterion
+
+```rust
+use criterion::{black_box, criterion_group, criterion_main, Criterion};
+
+fn expensive_operation(n: u64) -> u64 {
+    // Operation to benchmark
+    (0..n).sum()
+}
+
+fn bench_operation(c: &mut Criterion) {
+    c.bench_function("expensive_operation", |b| {
+        b.iter(|| expensive_operation(black_box(1000)))
+    });
+}
+
+criterion_group!(benches, bench_operation);
+criterion_main!(benches);
+
+// Run: cargo bench
+```
+
+#### Profiling with Flamegraph
+
+```bash
+# Install flamegraph
+cargo install flamegraph
+
+# Profile and generate flamegraph
+cargo flamegraph --bin my_binary
+
+# Opens flamegraph.svg showing hot paths
+```
+
+## Common Performance Bottlenecks
+
+### 1. Database Queries (N+1 Problem)
+
+```python
+# Bad: N+1 queries
+def get_users_with_posts():
+    users = User.query.all()  # 1 query
+    for user in users:
+        posts = Post.query.filter_by(user_id=user.id).all()  # N queries!
+        user.posts = posts
+    return users
+
+# Good: Single query with join
+def get_users_with_posts():
+    return User.query.options(joinedload(User.posts)).all()  # 1 query
+```
+
+### 2. Unnecessary Loops
+
+```python
+# Bad: Nested loops O(n²)
+def find_duplicates(items):
+    duplicates = []
+    for i, item1 in enumerate(items):
+        for j, item2 in enumerate(items[i+1:]):
+            if item1 == item2:
+                duplicates.append(item1)
+    return duplicates
+
+# Good: Use set O(n)
+def find_duplicates(items):
+    seen = set()
+    duplicates = set()
+    for item in items:
+        if item in seen:
+            duplicates.add(item)
+        seen.add(item)
+    return list(duplicates)
+```
+
+### 3. Inefficient Data Structures
+
+```python
+# Bad: List lookups O(n)
+def process_items(items, allowed_ids):
+    allowed_ids_list = [1, 2, 3, ...]  # List
+    result = []
+    for item in items:
+        if item.id in allowed_ids_list:  # O(n) lookup
+            result.append(item)
+    return result
+
+# Good: Set lookups O(1)
+def process_items(items, allowed_ids):
+    allowed_ids_set = {1, 2, 3, ...}  # Set
+    result = []
+    for item in items:
+        if item.id in allowed_ids_set:  # O(1) lookup
+            result.append(item)
+    return result
+```
+
+### 4. Excessive Memory Allocation
+
+```python
+# Bad: Building string with concatenation
+def build_large_string(items):
+    result = ""
+    for item in items:
+        result += str(item) + "\n"  # Creates new string each time
+    return result
+
+# Good: Use list and join
+def build_large_string(items):
+    parts = []
+    for item in items:
+        parts.append(str(item))
+    return "\n".join(parts)  # Single allocation
+```
+
+### 5. Synchronous I/O
+
+```python
+# Bad: Sequential HTTP requests
+def fetch_all_users(user_ids):
+    users = []
+    for user_id in user_ids:
+        response = requests.get(f"/users/{user_id}")  # Blocks
+        users.append(response.json())
+    return users
+
+# Good: Concurrent requests
+async def fetch_all_users(user_ids):
+    async with aiohttp.ClientSession() as session:
+        tasks = [fetch_user(session, user_id) for user_id in user_ids]
+        return await asyncio.gather(*tasks)
+```
+
+## Performance Optimization Strategies
+
+### Strategy 1: Caching
+
+```python
+from functools import lru_cache
+
+@lru_cache(maxsize=128)
+def expensive_calculation(n):
+    # Expensive computation
+    return result
+
+# First call: computed
+result1 = expensive_calculation(10)
+
+# Second call: cached
+result2 = expensive_calculation(10)  # Instant
+```
+
+### Strategy 2: Lazy Loading
+
+```python
+class ExpensiveResource:
+    def __init__(self):
+        self._data = None
+
+    @property
+    def data(self):
+        if self._data is None:
+            self._data = load_expensive_data()  # Only load when needed
+        return self._data
+```
+
+### Strategy 3: Batching
+
+```python
+# Bad: Process one at a time
+for item in items:
+    process_single(item)  # Many round trips
+
+# Good: Process in batches
+for batch in chunks(items, batch_size=100):
+    process_batch(batch)  # Fewer round trips
+```
+
+### Strategy 4: Indexing (Databases)
+
+```sql
+-- Bad: Sequential scan
+SELECT * FROM users WHERE email = 'user@example.com';
+
+-- Good: Use index
+CREATE INDEX idx_users_email ON users(email);
+SELECT * FROM users WHERE email = 'user@example.com';
+```
+
+### Strategy 5: Algorithmic Optimization
+
+```python
+# Bad: Bubble sort O(n²)
+def sort_items(items):
+    for i in range(len(items)):
+        for j in range(len(items) - 1):
+            if items[j] > items[j + 1]:
+                items[j], items[j + 1] = items[j + 1], items[j]
+
+# Good: Use built-in sort O(n log n)
+def sort_items(items):
+    return sorted(items)
+```
+
+## Performance Testing
+
+### Load Testing
+
+```python
+# Using locust for load testing
+from locust import HttpUser, task, between
+
+class WebsiteUser(HttpUser):
+    wait_time = between(1, 5)
+
+    @task
+    def load_homepage(self):
+        self.client.get("/")
+
+    @task
+    def load_user_profile(self):
+        self.client.get("/profile/123")
+
+# Run: locust -f locustfile.py
+# Shows requests/sec, response times, failure rate
+```
+
+### Stress Testing
+
+```bash
+# Using Apache Bench
+ab -n 10000 -c 100 http://localhost:8000/
+
+# n: total requests
+# c: concurrent requests
+# Shows: requests/sec, time per request, transfer rate
+```
+
+## Performance Monitoring
+
+### Key Metrics to Track
+
+```
+CPU Usage:
+- % CPU time
+- CPU cores utilized
+- System vs user time
+
+Memory Usage:
+- Heap size
+- Peak memory
+- Memory leaks (growing over time)
+
+I/O:
+- Disk read/write IOPS
+- Network throughput
+- Database query time
+
+Response Time:
+- p50 (median)
+- p95 (95th percentile)
+- p99 (99th percentile)
+- Max response time
+```
+
+### Application Performance Monitoring (APM)
+
+```python
+# Example: Using OpenTelemetry
+from opentelemetry import trace
+from opentelemetry.sdk.trace import TracerProvider
+
+trace.set_tracer_provider(TracerProvider())
+tracer = trace.get_tracer(__name__)
+
+def process_request():
+    with tracer.start_as_current_span("process_request"):
+        # Your code here
+        with tracer.start_as_current_span("database_query"):
+            fetch_data()
+        with tracer.start_as_current_span("process_data"):
+            process_data()
+```
+
+## Optimization Anti-Patterns
+
+### ❌ Optimizing Before Profiling
+
+```python
+# Bad: Optimizing blindly
+# "This looks slow, let me make it faster"
+result = [complex_operation(x) for x in items]  # Is this actually slow?
+```
+
+### ❌ Micro-Optimizations
+
+```python
+# Bad: Premature micro-optimization
+# Optimizing: x = x + 1  vs  x += 1
+# Impact: negligible (nanoseconds)
+# Don't waste time on this without proof it matters
+```
+
+### ❌ Sacrificing Readability
+
+```python
+# Bad: Unreadable "optimized" code
+r=[x for x in d if x>0and x<100and x%2==0]  # What does this do?
+
+# Good: Readable code (optimize only if profiling shows need)
+even_numbers = [
+    num for num in data
+    if 0 < num < 100 and num % 2 == 0
+]
+```
+
+## Quick Profiling Checklist
+
+Before optimizing:
+
+```
+□ Have you profiled to identify bottlenecks?
+□ Have you established baseline metrics?
+□ Are you optimizing the critical path?
+□ Have you checked for N+1 queries?
+□ Have you looked at algorithm complexity?
+□ Have you considered caching?
+□ Have you tested with realistic data volumes?
+□ Have you measured the impact of changes?
+□ Does the optimization maintain correctness?
+□ Is the code still readable?
+```
+
+## Remember
+
+- **Profile first** - Don't guess where the bottleneck is
+- **80/20 rule** - 20% of code accounts for 80% of execution time
+- **Real data** - Profile with production-like data volumes
+- **Diminishing returns** - Getting 2x faster is easy, 10x is hard
+- **Correctness > Speed** - Fast but wrong is useless