adversarial-workflow 0.6.5__py3-none-any.whl → 0.7.0__py3-none-any.whl

adversarial_workflow/__init__.py

@@ -12,7 +12,7 @@ Usage:
     adversarial validate "pytest"
 """
 
-__version__ = "0.6.5"
+__version__ = "0.7.0"
 __author__ = "Fredrik Matheson"
 __license__ = "MIT"
 

adversarial_workflow/cli.py

@@ -13,6 +13,7 @@ Commands:
     review - Run Phase 3: Code review
     validate - Run Phase 4: Test validation
     split - Split large task files into smaller evaluable chunks
+    check-citations - Verify URLs in documents before evaluation
 """
 
 import argparse
@@ -29,7 +30,7 @@ from typing import Dict, List, Optional, Tuple
 import yaml
 from dotenv import dotenv_values, load_dotenv
 
-__version__ = "0.6.5"
+__version__ = "0.7.0"
 
 # ANSI color codes for better output
 RESET = "\033[0m"
@@ -2819,6 +2820,106 @@ def list_evaluators() -> int:
     return 0
 
 
+def check_citations(
+    file_path: str,
+    output_tasks: Optional[str] = None,
+    mark_inline: bool = False,
+    concurrency: int = 10,
+    timeout: int = 10,
+) -> int:
+    """
+    Check citations (URLs) in a document.
+
+    Args:
+        file_path: Path to document to check
+        output_tasks: Optional path to write blocked URL tasks
+        mark_inline: Whether to mark URLs inline with status badges
+        concurrency: Maximum concurrent URL checks
+        timeout: Timeout per URL in seconds
+
+    Returns:
+        0 on success, 1 on error
+    """
+    from adversarial_workflow.utils.citations import (
+        URLStatus,
+        check_urls,
+        extract_urls,
+        generate_blocked_tasks,
+        mark_urls_inline,
+        print_verification_summary,
+    )
+
+    # Check file exists
+    if not os.path.exists(file_path):
+        print(f"{RED}Error: File not found: {file_path}{RESET}")
+        return 1
+
+    # Validate parameters
+    if concurrency < 1:
+        print(f"{RED}Error: Concurrency must be at least 1, got {concurrency}{RESET}")
+        return 1
+    if timeout < 1:
+        print(f"{RED}Error: Timeout must be at least 1 second, got {timeout}{RESET}")
+        return 1
+
+    print(f"🔗 Checking citations in: {file_path}")
+    print()
+
+    # Read document
+    with open(file_path, encoding="utf-8") as f:
+        document = f.read()
+
+    # Extract URLs
+    extracted = extract_urls(document)
+    urls = [e.url for e in extracted]
+
+    if not urls:
+        print(f"{YELLOW}No URLs found in document.{RESET}")
+        return 0
+
+    print(f"   Found {len(urls)} URLs to check")
+    print(f"   Checking with concurrency={concurrency}, timeout={timeout}s...")
+    print()
+
+    # Check URLs
+    results = check_urls(
+        urls,
+        concurrency=concurrency,
+        timeout=timeout,
+    )
+
+    # Print summary
+    print_verification_summary(results)
+
+    # Count blocked/broken
+    blocked_count = sum(1 for r in results if r.status in (URLStatus.BLOCKED, URLStatus.BROKEN))
+
+    # Mark document inline if requested
+    if mark_inline and results:
+        marked_document = mark_urls_inline(document, results)
+        if marked_document != document:
+            with open(file_path, "w", encoding="utf-8") as f:
+                f.write(marked_document)
+            print("\n   ✅ Updated document with status badges")
+
+    # Generate blocked tasks if requested or if there are blocked URLs
+    if blocked_count > 0:
+        if output_tasks:
+            output_path = Path(output_tasks)
+        else:
+            # Default to .adversarial/blocked-citations/
+            output_dir = Path.cwd() / ".adversarial" / "blocked-citations"
+            output_dir.mkdir(parents=True, exist_ok=True)
+            base_name = Path(file_path).stem
+            output_path = output_dir / f"{base_name}-blocked-urls.md"
+
+        task_content = generate_blocked_tasks(results, file_path, output_path)
+        if task_content:
+            print(f"   📋 Blocked URL tasks: {output_path}")
+
+    return 0
+
+
 def main():
     """Main CLI entry point."""
     import logging
@@ -2862,6 +2963,7 @@ def main():
         "validate",
         "review",
         "list-evaluators",
+        "check-citations",
     }
 
     parser = argparse.ArgumentParser(
@@ -2879,6 +2981,7 @@ Examples:
   adversarial review                    # Review implementation
   adversarial validate "npm test"       # Validate with tests
   adversarial split large-task.md       # Split large files
+  adversarial check-citations doc.md    # Verify URLs in document
 
 For more information: https://github.com/movito/adversarial-workflow
         """,
@@ -2961,6 +3064,38 @@ For more information: https://github.com/movito/adversarial-workflow
         help="List all available evaluators (built-in and local)",
     )
 
+    # check-citations command
+    citations_parser = subparsers.add_parser(
+        "check-citations",
+        help="Verify URLs in a document before evaluation",
+    )
+    citations_parser.add_argument("file", help="Document to check citations in")
+    citations_parser.add_argument(
+        "--output-tasks",
+        "-o",
+        help="Output file for blocked URL tasks (markdown)",
+    )
+    citations_parser.add_argument(
+        "--mark-inline",
+        action="store_true",
+        default=False,
+        help="Mark URLs inline with status badges (modifies document)",
+    )
+    citations_parser.add_argument(
+        "--concurrency",
+        "-c",
+        type=int,
+        default=10,
+        help="Maximum concurrent URL checks (default: 10)",
+    )
+    citations_parser.add_argument(
+        "--timeout",
+        "-t",
+        type=int,
+        default=10,
+        help="Timeout per URL in seconds (default: 10)",
+    )
+
     # Dynamic evaluator registration
     try:
         evaluators = get_all_evaluators()
@@ -3009,6 +3144,11 @@ For more information: https://github.com/movito/adversarial-workflow
             default=None,
             help="Timeout in seconds (default: from evaluator config or 180, max: 600)",
         )
+        eval_parser.add_argument(
+            "--check-citations",
+            action="store_true",
+            help="Verify URLs in document before evaluation",
+        )
         # Store config for later execution
         eval_parser.set_defaults(evaluator_config=config)
 
@@ -3044,6 +3184,16 @@ For more information: https://github.com/movito/adversarial-workflow
         # Log actual timeout and source
         print(f"Using timeout: {timeout}s ({source})")
 
+        # Check citations first if requested (read-only, doesn't modify file)
+        if getattr(args, "check_citations", False):
+            print()
+            result = check_citations(args.file, mark_inline=False)
+            if result != 0:
+                print(
+                    f"{YELLOW}Warning: Citation check had issues, continuing with evaluation...{RESET}"
+                )
+            print()
+
         return run_evaluator(
             args.evaluator_config,
             args.file,
@@ -3083,6 +3233,14 @@ For more information: https://github.com/movito/adversarial-workflow
         )
     elif args.command == "list-evaluators":
         return list_evaluators()
+    elif args.command == "check-citations":
+        return check_citations(
+            args.file,
+            output_tasks=args.output_tasks,
+            mark_inline=args.mark_inline,
+            concurrency=args.concurrency,
+            timeout=args.timeout,
+        )
     else:
         parser.print_help()
         return 1

adversarial_workflow/evaluators/runner.py

@@ -139,6 +139,7 @@ def _run_custom_evaluator(
             "--model",
             config.model,
             "--yes",
+            "--no-detect-urls",
             "--no-git",
             "--no-auto-commits",
             "--message-file",

adversarial_workflow/utils/citations.py

@@ -0,0 +1,643 @@
+"""
+Citation verification utilities for checking URLs in documents.
+
+This module provides:
+- URL extraction from markdown documents
+- Async parallel URL checking with caching
+- Inline marking of URL status
+- Blocked URL task file generation
+
+Status categories:
+- available: 200 OK, content accessible
+- blocked: Paywall/auth/bot-blocked (401, 403, or bot detection)
+- broken: 404, 500, timeout, DNS failure
+- redirect: 301/302 with final destination noted
+"""
+
+import asyncio
+import hashlib
+import json
+import logging
+import os
+import re
+import time
+from dataclasses import dataclass
+from datetime import datetime, timezone
+from enum import Enum
+from pathlib import Path
+from typing import Optional
+
+# Module logger for debugging URL check failures
+logger = logging.getLogger(__name__)
+
+
+class URLStatus(Enum):
+    """URL verification status categories."""
+
+    AVAILABLE = "available"
+    BLOCKED = "blocked"
+    BROKEN = "broken"
+    REDIRECT = "redirect"
+
+
+@dataclass
+class URLResult:
+    """Result of checking a single URL."""
+
+    url: str
+    status: URLStatus
+    status_code: Optional[int] = None
+    final_url: Optional[str] = None
+    error: Optional[str] = None
+    checked_at: Optional[float] = None
+
+    def to_dict(self) -> dict:
+        """Convert to dictionary for JSON serialization."""
+        return {
+            "url": self.url,
+            "status": self.status.value,
+            "status_code": self.status_code,
+            "final_url": self.final_url,
+            "error": self.error,
+            "checked_at": self.checked_at,
+        }
+
+    @classmethod
+    def from_dict(cls, data: dict) -> "URLResult":
+        """Create from dictionary."""
+        return cls(
+            url=data["url"],
+            status=URLStatus(data["status"]),
+            status_code=data.get("status_code"),
+            final_url=data.get("final_url"),
+            error=data.get("error"),
+            checked_at=data.get("checked_at"),
+        )
+
+
+@dataclass
+class ExtractedURL:
+    """A URL extracted from a document with context."""
+
+    url: str
+    position: int
+    context: str
+    line_number: int
+
+
+# URL extraction pattern - matches http/https URLs
+URL_PATTERN = re.compile(r"https?://[^\s\)\]\>\"\'\`]+")
+
+# Bot detection patterns in response
+BOT_DETECTION_PATTERNS = [
+    "captcha",
+    "cloudflare",
+    "access denied",
+    "forbidden",
+    "bot detected",
+    "please verify",
+    "human verification",
+]
+
+# Default configuration
+DEFAULT_CONFIG = {
+    "max_urls": 100,
+    "concurrency": 10,
+    "timeout_per_url": 10,
+    "cache_ttl": 86400,  # 24 hours
+}
+
+
+def extract_urls(document: str, max_urls: int = 100) -> list[ExtractedURL]:
+    """
+    Extract URLs from a document with surrounding context.
+
+    Args:
+        document: The document text to extract URLs from
+        max_urls: Maximum number of URLs to extract (default: 100)
+
+    Returns:
+        List of ExtractedURL objects with position and context
+    """
+    urls = []
+    lines = document.split("\n")
+    line_starts = [0]
+    for line in lines:
+        line_starts.append(line_starts[-1] + len(line) + 1)
+
+    for match in URL_PATTERN.finditer(document):
+        url = match.group().rstrip(".,;:!?")  # Clean trailing punctuation
+        position = match.start()
+
+        # Find line number
+        line_number = 1
+        for i, start in enumerate(line_starts):
+            if start > position:
+                line_number = i
+                break
+
+        # Get context (50 chars before and after)
+        start = max(0, position - 50)
+        end = min(len(document), match.end() + 50)
+        context = document[start:end]
+
+        urls.append(
+            ExtractedURL(
+                url=url,
+                position=position,
+                context=context,
+                line_number=line_number,
+            )
+        )
+
+        if len(urls) >= max_urls:
+            break
+
+    return urls
+
+
+def get_cache_path(cache_dir: Optional[Path] = None) -> Path:
+    """Get the path to the URL cache file."""
+    if cache_dir is None:
+        cache_dir = Path.cwd() / ".adversarial"
+    cache_dir.mkdir(parents=True, exist_ok=True)
+    return cache_dir / "url_cache.json"
+
+
+def load_cache(cache_path: Path) -> dict[str, dict]:
+    """Load URL cache from disk."""
+    if not cache_path.exists():
+        return {}
+    try:
+        with open(cache_path) as f:
+            return json.load(f)
+    except (json.JSONDecodeError, OSError):
+        return {}
+
+
+def save_cache(cache_path: Path, cache: dict[str, dict]) -> None:
+    """Save URL cache to disk."""
+    with open(cache_path, "w") as f:
+        json.dump(cache, f, indent=2)
+
+
+def get_cache_key(url: str) -> str:
+    """Generate a cache key for a URL."""
+    return hashlib.md5(url.encode()).hexdigest()
+
+
+def classify_response(status_code: int, _headers: dict, content: Optional[str] = None) -> URLStatus:
+    """
+    Classify HTTP response into a URL status.
+
+    Args:
+        status_code: HTTP status code
+        _headers: Response headers (reserved for future use)
+        content: Optional response body content (for bot detection)
+
+    Returns:
+        URLStatus enum value
+    """
+    if status_code == 200:
+        # Check for bot blocking in content
+        if content:
+            content_lower = content.lower()
+            for pattern in BOT_DETECTION_PATTERNS:
+                if pattern in content_lower:
+                    return URLStatus.BLOCKED
+        return URLStatus.AVAILABLE
+    elif status_code in (301, 302, 307, 308):
+        return URLStatus.REDIRECT
+    elif status_code in (401, 403):
+        return URLStatus.BLOCKED
+    elif status_code == 429:
+        return URLStatus.BLOCKED  # Rate limited
+    else:
+        return URLStatus.BROKEN
+
+
+async def check_url_async(
+    url: str,
+    timeout: int = 10,
+    session=None,
+) -> URLResult:
+    """
+    Check a single URL asynchronously.
+
+    Args:
+        url: URL to check
+        timeout: Request timeout in seconds
+        session: Optional aiohttp session to reuse
+
+    Returns:
+        URLResult with status information
+    """
+    try:
+        import aiohttp
+    except ImportError:
+        return URLResult(
+            url=url,
+            status=URLStatus.BROKEN,
+            error="aiohttp not installed - run: pip install aiohttp",
+            checked_at=time.time(),
+        )
+
+    close_session = False
+    if session is None:
+        session = aiohttp.ClientSession()
+        close_session = True
+
+    try:
+        async with session.head(
+            url,
+            timeout=aiohttp.ClientTimeout(total=timeout),
+            allow_redirects=True,
+            headers={"User-Agent": "Mozilla/5.0 (compatible; CitationVerifier/1.0)"},
+        ) as response:
+            final_url = str(response.url) if str(response.url) != url else None
+            status = classify_response(response.status, dict(response.headers))
+
+            # If redirect to an available page, mark as redirect (informational)
+            # Keep broken/blocked status if redirect leads to error page
+            if final_url and response.history and status == URLStatus.AVAILABLE:
+                status = URLStatus.REDIRECT
+
+            return URLResult(
+                url=url,
+                status=status,
+                status_code=response.status,
+                final_url=final_url,
+                checked_at=time.time(),
+            )
+    except asyncio.TimeoutError:
+        return URLResult(
+            url=url,
+            status=URLStatus.BROKEN,
+            error="Timeout",
+            checked_at=time.time(),
+        )
+    except Exception as e:
+        error_name = type(e).__name__
+        # Log full exception for debugging while returning truncated message
+        logger.debug("URL check failed for %s: %s", url, e, exc_info=True)
+        return URLResult(
+            url=url,
+            status=URLStatus.BROKEN,
+            error=f"{error_name}: {str(e)[:50]}",
+            checked_at=time.time(),
+        )
+    finally:
+        if close_session:
+            await session.close()
+
+
+async def check_urls_parallel(
+    urls: list[str],
+    concurrency: int = 10,
+    timeout: int = 10,
+    cache: Optional[dict] = None,
+    cache_ttl: int = 86400,
+) -> list[URLResult]:
+    """
+    Check multiple URLs in parallel with optional caching.
+
+    Args:
+        urls: List of URLs to check
+        concurrency: Maximum concurrent requests (must be >= 1)
+        timeout: Timeout per request in seconds (must be >= 1)
+        cache: Optional cache dictionary
+        cache_ttl: Cache TTL in seconds (default: 24 hours)
+
+    Returns:
+        List of URLResult objects
+
+    Raises:
+        ValueError: If concurrency or timeout is less than 1
+    """
+    # Validate parameters to prevent deadlocks
+    if concurrency < 1:
+        raise ValueError(f"concurrency must be >= 1, got {concurrency}")
+    if timeout < 1:
+        raise ValueError(f"timeout must be >= 1, got {timeout}")
+
+    try:
+        import aiohttp
+    except ImportError:
+        return [
+            URLResult(
+                url=url,
+                status=URLStatus.BROKEN,
+                error="aiohttp not installed",
+                checked_at=time.time(),
+            )
+            for url in urls
+        ]
+
+    url_to_result: dict[str, URLResult] = {}
+    urls_to_check = []
+    current_time = time.time()
+
+    # Check cache first
+    if cache is not None:
+        for url in urls:
+            cache_key = get_cache_key(url)
+            if cache_key in cache:
+                cached = cache[cache_key]
+                if cached.get("expires", 0) > current_time:
+                    url_to_result[url] = URLResult.from_dict(cached["result"])
+                    continue
+            urls_to_check.append(url)
+    else:
+        urls_to_check = list(urls)
+
+    if urls_to_check:
+        # Create semaphore for concurrency limiting
+        semaphore = asyncio.Semaphore(concurrency)
+
+        async def check_with_semaphore(session, url):
+            async with semaphore:
+                return await check_url_async(url, timeout, session)
+
+        # Check remaining URLs
+        connector = aiohttp.TCPConnector(limit=concurrency, limit_per_host=5)
+        async with aiohttp.ClientSession(connector=connector) as session:
+            tasks = [check_with_semaphore(session, url) for url in urls_to_check]
+            checked_results = await asyncio.gather(*tasks)
+
+        # Update cache and store results
+        for result in checked_results:
+            if cache is not None:
+                cache_key = get_cache_key(result.url)
+                cache[cache_key] = {
+                    "result": result.to_dict(),
+                    "expires": current_time + cache_ttl,
+                }
+            url_to_result[result.url] = result
+
+    # Return results in original URL order
+    return [url_to_result[url] for url in urls]
+
+
+def check_urls(
+    urls: list[str],
+    concurrency: int = 10,
+    timeout: int = 10,
+    cache_dir: Optional[Path] = None,
+    cache_ttl: int = 86400,
+) -> list[URLResult]:
+    """
+    Check multiple URLs synchronously (wrapper around async version).
+
+    Args:
+        urls: List of URLs to check
+        concurrency: Maximum concurrent requests
+        timeout: Timeout per request in seconds
+        cache_dir: Optional cache directory
+        cache_ttl: Cache TTL in seconds
+
+    Returns:
+        List of URLResult objects
+
+    Raises:
+        RuntimeError: If called from within an async context (event loop running).
+            Use check_urls_parallel() directly from async code.
+    """
+    # Guard against calling from async context
+    try:
+        asyncio.get_running_loop()
+        raise RuntimeError(
+            "check_urls() cannot be called from within an async context. "
+            "Use check_urls_parallel() directly instead."
+        )
+    except RuntimeError as e:
+        # No running loop - this is expected, proceed
+        if "no running event loop" not in str(e).lower():
+            raise
+
+    # Load cache
+    cache_path = get_cache_path(cache_dir)
+    cache = load_cache(cache_path)
+
+    # Run async check
+    results = asyncio.run(
+        check_urls_parallel(
+            urls,
+            concurrency=concurrency,
+            timeout=timeout,
+            cache=cache,
+            cache_ttl=cache_ttl,
+        )
+    )
+
+    # Save cache
+    save_cache(cache_path, cache)
+
+    return results
+
+
+def get_status_badge(result: URLResult) -> str:
+    """
+    Generate an inline status badge for a URL result.
+
+    Args:
+        result: URLResult to generate badge for
+
+    Returns:
+        Markdown-formatted status badge
+    """
+    if result.status == URLStatus.AVAILABLE:
+        return f"[✅ Verified | {result.status_code} OK]"
+    elif result.status == URLStatus.BLOCKED:
+        if result.status_code:
+            return f"[⚠️ Blocked | {result.status_code}]"
+        return "[⚠️ Blocked | Access Denied]"
+    elif result.status == URLStatus.BROKEN:
+        if result.error:
+            return f"[❌ Broken | {result.error}]"
+        if result.status_code:
+            return f"[❌ Broken | {result.status_code}]"
+        return "[❌ Broken | Unreachable]"
+    elif result.status == URLStatus.REDIRECT:
+        dest = (
+            result.final_url[:30] + "..."
+            if result.final_url and len(result.final_url) > 30
+            else result.final_url
+        )
+        return f"[🔄 Redirect | → {dest}]"
+    return "[❓ Unknown]"
+
+
+def mark_urls_inline(document: str, results: list[URLResult]) -> str:
+    """
+    Mark URLs in a document with their status badges.
+
+    Args:
+        document: Original document text
+        results: List of URL check results
+
+    Returns:
+        Document with inline status badges added after URLs
+    """
+    # Create URL to result mapping
+    url_results = {r.url: r for r in results}
+
+    # Find all URLs and their positions
+    marked = document
+    offset = 0  # Track offset as we insert badges
+
+    for match in URL_PATTERN.finditer(document):
+        url = match.group().rstrip(".,;:!?")  # Same stripping as extract_urls
+        if url in url_results:
+            result = url_results[url]
+            badge = get_status_badge(result)
+
+            # Check if badge already exists after this URL
+            end_pos = match.end() + offset
+            remaining = marked[end_pos:]
+            if remaining.startswith((" [✅", " [⚠️", " [❌", " [🔄")):
+                continue  # Already marked
+
+            # Insert badge after URL
+            insert_pos = end_pos
+            marked = marked[:insert_pos] + " " + badge + marked[insert_pos:]
+            offset += len(badge) + 1
+
+    return marked
+
+
+def generate_blocked_tasks(
+    results: list[URLResult],
+    document_path: str,
+    output_path: Optional[Path] = None,
+) -> str:
+    """
+    Generate a task file for blocked URLs requiring manual verification.
+
+    Args:
+        results: List of URL check results
+        document_path: Path to the source document
+        output_path: Optional path to write task file
+
+    Returns:
+        Task file content as string
+    """
+    blocked = [r for r in results if r.status in (URLStatus.BLOCKED, URLStatus.BROKEN)]
+
+    if not blocked:
+        return ""
+
+    timestamp = datetime.now(timezone.utc).strftime("%Y-%m-%d %H:%M UTC")
+    content = f"""# Blocked Citation Verification Tasks
+
+**Source**: {document_path}
+**Generated**: {timestamp}
+**Total blocked URLs**: {len(blocked)}
+
+## URLs Requiring Manual Verification
+
+"""
+
+    for i, result in enumerate(blocked, 1):
+        status_label = "⚠️ Blocked" if result.status == URLStatus.BLOCKED else "❌ Broken"
+        reason = result.error or (f"HTTP {result.status_code}" if result.status_code else "Unknown")
+
+        content += f"""### {i}. {status_label}
+
+- **URL**: {result.url}
+- **Reason**: {reason}
+- [ ] Verify URL manually
+- [ ] Update document if URL is permanently unavailable
+
+"""
+
+    content += """---
+
+## Instructions
+
+1. Open each URL in a browser
+2. Verify if content is accessible
+3. If blocked by paywall/auth, note the access method needed
+4. If broken, find replacement URL or remove citation
+5. Update the source document accordingly
+"""
+
+    if output_path:
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+        with open(output_path, "w", encoding="utf-8") as f:
+            f.write(content)
+
+    return content
+
+
+def verify_document(
+    document_path: Path,
+    output_tasks_path: Optional[Path] = None,
+    mark_inline: bool = True,
+    concurrency: int = 10,
+    timeout: int = 10,
+    cache_dir: Optional[Path] = None,
+) -> tuple[str, list[URLResult], str]:
+    """
+    Verify all citations in a document.
+
+    Args:
+        document_path: Path to the document to verify
+        output_tasks_path: Optional path for blocked URL task file
+        mark_inline: Whether to mark URLs inline in the document
+        concurrency: Maximum concurrent requests
+        timeout: Timeout per request
+        cache_dir: Optional cache directory
+
+    Returns:
+        Tuple of (marked_document, results, blocked_tasks)
+    """
+    with open(document_path, encoding="utf-8") as f:
+        document = f.read()
+
+    # Extract URLs
+    extracted = extract_urls(document)
+    urls = [e.url for e in extracted]
+
+    if not urls:
+        return document, [], ""
+
+    # Check URLs
+    results = check_urls(
+        urls,
+        concurrency=concurrency,
+        timeout=timeout,
+        cache_dir=cache_dir,
+    )
+
+    # Mark document if requested
+    marked_document = document
+    if mark_inline:
+        marked_document = mark_urls_inline(document, results)
+
+    # Generate blocked tasks
+    blocked_tasks = generate_blocked_tasks(
+        results,
+        str(document_path),
+        output_tasks_path,
+    )
+
+    return marked_document, results, blocked_tasks
+
+
+def print_verification_summary(results: list[URLResult]) -> None:
+    """Print a summary of verification results to stdout."""
+    available = sum(1 for r in results if r.status == URLStatus.AVAILABLE)
+    blocked = sum(1 for r in results if r.status == URLStatus.BLOCKED)
+    broken = sum(1 for r in results if r.status == URLStatus.BROKEN)
+    redirect = sum(1 for r in results if r.status == URLStatus.REDIRECT)
+
+    total = len(results)
+    print("\n📋 Citation Verification Summary")
+    print(f"   Total URLs checked: {total}")
+    print(f"   ✅ Available: {available}")
+    print(f"   🔄 Redirect: {redirect}")
+    print(f"   ⚠️  Blocked: {blocked}")
+    print(f"   ❌ Broken: {broken}")
+
+    if blocked + broken > 0:
+        print(f"\n   ⚠️  {blocked + broken} URLs need manual verification")

{adversarial_workflow-0.6.5.dist-info → adversarial_workflow-0.7.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: adversarial-workflow
-Version: 0.6.5
+Version: 0.7.0
 Summary: Multi-stage AI evaluation system for task plans, code review, and test validation
 Author: Fredrik Matheson
 License: MIT
@@ -24,9 +24,11 @@ License-File: LICENSE
 Requires-Dist: pyyaml>=6.0
 Requires-Dist: python-dotenv>=0.19.0
 Requires-Dist: aider-chat>=0.86.0
+Requires-Dist: aiohttp>=3.8.0
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 Requires-Dist: pytest-cov>=3.0; extra == "dev"
+Requires-Dist: pytest-asyncio>=0.21.0; extra == "dev"
 Requires-Dist: black>=22.0; extra == "dev"
 Requires-Dist: isort>=5.0; extra == "dev"
 Requires-Dist: flake8>=4.0; extra == "dev"

{adversarial_workflow-0.6.5.dist-info → adversarial_workflow-0.7.0.dist-info}/RECORD

@@ -1,11 +1,11 @@
-adversarial_workflow/__init__.py,sha256=5yx_5uxdFrtwW3Ma6Z6SNPWM5SxjIcnmIggqgtpOzFE,596
+adversarial_workflow/__init__.py,sha256=Aj_FdCEOJYpaeOe9SMO1vWqlpzoR9AONRz2SZ8RKgVg,596
 adversarial_workflow/__main__.py,sha256=iM2jmO5YCFpGxfWiEhIYi_SsxVa0hRIE-MB7J0EcN7Y,120
-adversarial_workflow/cli.py,sha256=rQfym1MAvkfoefvzhUmxIFrohHTSxP9fR00ORpiXc4M,110577
+adversarial_workflow/cli.py,sha256=mKvOe3Q-afCfzXALNCQ47GBOSwccE17kqYpxOHbOC6k,115541
 adversarial_workflow/evaluators/__init__.py,sha256=A9ZKUmjSMfyvEu6jDzYAFLxfkt_OQ4RGA10Bv_eO2i4,1267
 adversarial_workflow/evaluators/builtins.py,sha256=u5LokYLe8ruEW2tunhOQaNSkpcZ9Ee2IeTkaC0dZDSY,1102
 adversarial_workflow/evaluators/config.py,sha256=H_4vkto07rAqnz0qEYdzN_DH6WbvRPMIEdkEOFE58UI,1651
 adversarial_workflow/evaluators/discovery.py,sha256=Fe_mj7S4aIzXuib28b_uLu3vH8AMAy1B3MDyssi3QU0,7953
-adversarial_workflow/evaluators/runner.py,sha256=d1Plh8iB2Ip7dtha6ULxjbB2vodVz1Z9kbWeMn7YjZg,9280
+adversarial_workflow/evaluators/runner.py,sha256=7zui3ec190pBC4VB3HRxVgexJJff5F2V6oxgJiMv2tc,9312
 adversarial_workflow/templates/.aider.conf.yml.template,sha256=jT2jWIgsnmS3HLhoQWMTO3GV07bUcsT2keYw60jqiDw,183
 adversarial_workflow/templates/.env.example.template,sha256=TmTlcgz44uZqIbqgXqdfHMl-0vVn96F_EGNohClFkb8,1821
 adversarial_workflow/templates/README.template,sha256=FQAMPO99eIt_kgQfwhGHcrK736rm_MEvWSbPnqBSjAE,1349
@@ -21,13 +21,14 @@ adversarial_workflow/templates/agent-context/agent-handoffs-minimal.json.templat
 adversarial_workflow/templates/agent-context/agent-handoffs.json.template,sha256=rdRX79xdNKyTGd8_g6pvSdQK5VC5sDiErq_OZFslSXI,3904
 adversarial_workflow/templates/agent-context/current-state.json.template,sha256=UH3SQGjXGNzScqHnQcPrhrI9ZvjQBC3oBp4S9TilzoY,2325
 adversarial_workflow/utils/__init__.py,sha256=Pnm-a_jqoMVOxHdvVWXeVrL0IKI-zkY7EAdbQmZAkSI,352
+adversarial_workflow/utils/citations.py,sha256=Xhla-M3az4aXtd9F0UHyBXhTBDyYOo9Mb1IUi8nPLPs,18619
 adversarial_workflow/utils/colors.py,sha256=uRrG6KfIDBLo0F5_vPwms9NCm9-x8YXBiyZ4naCr868,160
 adversarial_workflow/utils/config.py,sha256=3VmF65ItUbFzbyAZ1RUoOtpS_t6n1wqIhKft8eSNsdw,1303
 adversarial_workflow/utils/file_splitter.py,sha256=kvWh0xVjd08fsEXgysoHd5zFwJHqs-JRKottO8scYCA,12381
 adversarial_workflow/utils/validation.py,sha256=0QfuRd-kurcadUCd9XQvO-N8RsmLp6ONQnc0vaQTUBA,2188
-adversarial_workflow-0.6.5.dist-info/licenses/LICENSE,sha256=M-dOQlre-NmicyPa55hYOJUW8roGpCKEgtq-z0z1KCA,1073
-adversarial_workflow-0.6.5.dist-info/METADATA,sha256=1thtTKRADK52Y6EQLS8lwrhSwpPnkL8P2xumQBhvUjQ,30832
-adversarial_workflow-0.6.5.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
-adversarial_workflow-0.6.5.dist-info/entry_points.txt,sha256=9H-iZ-yF1uKZ8P0G1suc6kWR0NvK7uPZJbhN7nvt1sE,62
-adversarial_workflow-0.6.5.dist-info/top_level.txt,sha256=8irutNxLRjUbTlzfAibIpz7_ovkkF2h8ES69NQpv24c,21
-adversarial_workflow-0.6.5.dist-info/RECORD,,
+adversarial_workflow-0.7.0.dist-info/licenses/LICENSE,sha256=M-dOQlre-NmicyPa55hYOJUW8roGpCKEgtq-z0z1KCA,1073
+adversarial_workflow-0.7.0.dist-info/METADATA,sha256=s0lrhtLRaXy6HE-QeCAcmdH8mswaeyxTvQ1H3khhG1k,30916
+adversarial_workflow-0.7.0.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
+adversarial_workflow-0.7.0.dist-info/entry_points.txt,sha256=9H-iZ-yF1uKZ8P0G1suc6kWR0NvK7uPZJbhN7nvt1sE,62
+adversarial_workflow-0.7.0.dist-info/top_level.txt,sha256=8irutNxLRjUbTlzfAibIpz7_ovkkF2h8ES69NQpv24c,21
+adversarial_workflow-0.7.0.dist-info/RECORD,,