adversarial-workflow 0.6.3__tar.gz → 0.6.4__tar.gz

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: adversarial-workflow
-Version: 0.6.3
-Summary: Multi-stage AI code review system preventing phantom work - Author/Evaluator pattern
+Version: 0.6.4
+Summary: Multi-stage AI evaluation system for task plans, code review, and test validation
 Author: Fredrik Matheson
 License: MIT
 Project-URL: Homepage, https://github.com/movito/adversarial-workflow

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/adversarial_workflow/__init__.py

@@ -12,7 +12,7 @@ Usage:
     adversarial validate "pytest"
 """
 
-__version__ = "0.6.2"
+__version__ = "0.6.4"
 __author__ = "Fredrik Matheson"
 __license__ = "MIT"
 

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/adversarial_workflow/cli.py

@@ -29,7 +29,7 @@ from typing import Dict, List, Optional, Tuple
 import yaml
 from dotenv import dotenv_values, load_dotenv
 
-__version__ = "0.6.2"
+__version__ = "0.6.4"
 
 # ANSI color codes for better output
 RESET = "\033[0m"
@@ -180,7 +180,9 @@ def create_env_file_interactive(
     env_content += "# DO NOT COMMIT THIS FILE\n\n"
 
     if anthropic_key:
-        env_content += f"# Anthropic API Key (Claude 3.5 Sonnet)\nANTHROPIC_API_KEY={anthropic_key}\n\n"
+        env_content += (
+            f"# Anthropic API Key (Claude 3.5 Sonnet)\nANTHROPIC_API_KEY={anthropic_key}\n\n"
+        )
 
     if openai_key:
         env_content += f"# OpenAI API Key (GPT-4o)\nOPENAI_API_KEY={openai_key}\n\n"
@@ -255,9 +257,7 @@ def init_interactive(project_path: str = ".") -> int:
             ],
         )
 
-        anthropic_key = prompt_user(
-            "Paste your Anthropic API key (or Enter to skip)", secret=True
-        )
+        anthropic_key = prompt_user("Paste your Anthropic API key (or Enter to skip)", secret=True)
 
         if anthropic_key:
             is_valid, message = validate_api_key(anthropic_key, "anthropic")
@@ -281,9 +281,7 @@ def init_interactive(project_path: str = ".") -> int:
             ],
         )
 
-        openai_key = prompt_user(
-            "Paste your OpenAI API key (or Enter to skip)", secret=True
-        )
+        openai_key = prompt_user("Paste your OpenAI API key (or Enter to skip)", secret=True)
 
         if openai_key:
             is_valid, message = validate_api_key(openai_key, "openai")
@@ -613,9 +611,7 @@ def init(project_path: str = ".", interactive: bool = True) -> int:
         print("   3. Then run: adversarial init")
         print()
         print(f"{BOLD}HELP:{RESET}")
-        print(
-            "   New to git? https://git-scm.com/book/en/v2/Getting-Started-Installing-Git"
-        )
+        print("   New to git? https://git-scm.com/book/en/v2/Getting-Started-Installing-Git")
         return 1
 
     # Pre-flight validation: Check package integrity
@@ -649,9 +645,7 @@ def init(project_path: str = ".", interactive: bool = True) -> int:
             print(f"   • {template}")
         print()
         print(f"{BOLD}FIX:{RESET}")
-        print(
-            "   1. Report this issue: https://github.com/movito/adversarial-workflow/issues"
-        )
+        print("   1. Report this issue: https://github.com/movito/adversarial-workflow/issues")
         print(
             "   2. Or try reinstalling: pip install --upgrade --force-reinstall adversarial-workflow"
         )
@@ -1030,13 +1024,9 @@ def check() -> int:
     else:
         status_parts = []
         if error_count > 0:
-            status_parts.append(
-                f"{error_count} error" + ("s" if error_count != 1 else "")
-            )
+            status_parts.append(f"{error_count} error" + ("s" if error_count != 1 else ""))
         if warning_count > 0:
-            status_parts.append(
-                f"{warning_count} warning" + ("s" if warning_count != 1 else "")
-            )
+            status_parts.append(f"{warning_count} warning" + ("s" if warning_count != 1 else ""))
         if info_count > 0:
             status_parts.append(f"{info_count} info")
 
@@ -1088,20 +1078,14 @@ def health(verbose: bool = False, json_output: bool = False) -> int:
     # Helper functions for tracking check results
     def check_pass(category: str, message: str, detail: str = None):
         nonlocal passed
-        results[category].append(
-            {"status": "pass", "message": message, "detail": detail}
-        )
+        results[category].append({"status": "pass", "message": message, "detail": detail})
         if not json_output:
             print(f"  {GREEN}✅{RESET} {message}")
         passed += 1
 
-    def check_warn(
-        category: str, message: str, detail: str = None, recommendation: str = None
-    ):
+    def check_warn(category: str, message: str, detail: str = None, recommendation: str = None):
         nonlocal warnings
-        results[category].append(
-            {"status": "warn", "message": message, "detail": detail}
-        )
+        results[category].append({"status": "warn", "message": message, "detail": detail})
         if not json_output:
             print(f"  {YELLOW}⚠️{RESET}  {message}")
             if detail and verbose:
@@ -1110,9 +1094,7 @@ def health(verbose: bool = False, json_output: bool = False) -> int:
             recommendations.append(recommendation)
         warnings += 1
 
-    def check_fail(
-        category: str, message: str, fix: str = None, recommendation: str = None
-    ):
+    def check_fail(category: str, message: str, fix: str = None, recommendation: str = None):
         nonlocal errors
         results[category].append({"status": "fail", "message": message, "fix": fix})
         if not json_output:
@@ -1124,9 +1106,7 @@ def health(verbose: bool = False, json_output: bool = False) -> int:
         errors += 1
 
     def check_info(category: str, message: str, detail: str = None):
-        results[category].append(
-            {"status": "info", "message": message, "detail": detail}
-        )
+        results[category].append({"status": "info", "message": message, "detail": detail})
         if not json_output:
             print(f"  {CYAN}ℹ️{RESET}  {message}")
             if detail and verbose:
@@ -1258,23 +1238,13 @@ def health(verbose: bool = False, json_output: bool = False) -> int:
                 )
                 if git_status.returncode == 0:
                     modified = len(
-                        [
-                            l
-                            for l in git_status.stdout.splitlines()
-                            if l.startswith(" M")
-                        ]
+                        [l for l in git_status.stdout.splitlines() if l.startswith(" M")]
                     )
                     untracked = len(
-                        [
-                            l
-                            for l in git_status.stdout.splitlines()
-                            if l.startswith("??")
-                        ]
+                        [l for l in git_status.stdout.splitlines() if l.startswith("??")]
                     )
                     if modified == 0 and untracked == 0:
-                        check_pass(
-                            "dependencies", f"Git: {version} (working tree clean)"
-                        )
+                        check_pass("dependencies", f"Git: {version} (working tree clean)")
                     else:
                         check_info(
                             "dependencies",
@@ -1311,11 +1281,7 @@ def health(verbose: bool = False, json_output: bool = False) -> int:
             aider_version = subprocess.run(
                 ["aider", "--version"], capture_output=True, text=True, timeout=2
             )
-            version = (
-                aider_version.stdout.strip()
-                if aider_version.returncode == 0
-                else "unknown"
-            )
+            version = aider_version.stdout.strip() if aider_version.returncode == 0 else "unknown"
             check_pass("dependencies", f"Aider: {version} (functional)")
         except:
             check_pass("dependencies", "Aider: installed")
@@ -1459,9 +1425,7 @@ def health(verbose: bool = False, json_output: bool = False) -> int:
                     json.load(f)
                 check_pass("agent_coordination", "current-state.json - Valid JSON")
             except json.JSONDecodeError as e:
-                check_fail(
-                    "agent_coordination", f"current-state.json - Invalid JSON: {e}"
-                )
+                check_fail("agent_coordination", f"current-state.json - Invalid JSON: {e}")
         else:
             check_info("agent_coordination", "current-state.json not found (optional)")
 
@@ -1505,9 +1469,7 @@ def health(verbose: bool = False, json_output: bool = False) -> int:
                     with open(script_path) as f:
                         content = f.read()
                     if "#!/bin/bash" in content or "#!/usr/bin/env bash" in content:
-                        check_pass(
-                            "workflow_scripts", f"{script_name} - Executable, valid"
-                        )
+                        check_pass("workflow_scripts", f"{script_name} - Executable, valid")
                     else:
                         check_warn(
                             "workflow_scripts",
@@ -1781,9 +1743,7 @@ def verify_token_count(task_file: str, log_file: str) -> None:
             f"   Difference: {expected_tokens - actual_tokens:,} tokens ({100 - int(actual_tokens/expected_tokens*100)}% less)"
         )
         print()
-        print(
-            f"{BOLD}Note:{RESET} Large files may not be fully processed by evaluator."
-        )
+        print(f"{BOLD}Note:{RESET} Large files may not be fully processed by evaluator.")
         print(f"      Consider splitting into smaller documents (<1,000 lines).")
         print()
 
@@ -1964,9 +1924,7 @@ def evaluate(task_file: str) -> int:
             print(f"{RED}❌ ERROR: OpenAI rate limit exceeded{RESET}")
             print()
             print(f"{BOLD}WHY:{RESET}")
-            print(
-                "   Your task file is too large for your OpenAI organization's rate limit"
-            )
+            print("   Your task file is too large for your OpenAI organization's rate limit")
             print()
 
             # Extract file size for helpful message
@@ -2013,9 +1971,7 @@ def evaluate(task_file: str) -> int:
             print()
             print(f"{BOLD}FIX:{RESET}")
             print("   Option 1 (RECOMMENDED): Use WSL (Windows Subsystem for Linux)")
-            print(
-                "     1. Install WSL: https://learn.microsoft.com/windows/wsl/install"
-            )
+            print("     1. Install WSL: https://learn.microsoft.com/windows/wsl/install")
             print("     2. Open WSL terminal")
             print("     3. Reinstall package in WSL: pip install adversarial-workflow")
             print()
@@ -2186,9 +2142,7 @@ def validate(test_command: Optional[str] = None) -> int:
         return 1
 
     try:
-        result = subprocess.run(
-            [script, test_command], timeout=600
-        )  # 10 minutes for tests
+        result = subprocess.run([script, test_command], timeout=600)  # 10 minutes for tests
     except subprocess.TimeoutExpired:
         print(f"{RED}❌ ERROR: Test validation timed out (>10 minutes){RESET}")
         return 1
@@ -2241,9 +2195,7 @@ def select_agent_template() -> Dict[str, str]:
     elif choice == "3":
         print()
         print(f"{CYAN}Custom Template URL:{RESET}")
-        print(
-            "  Example: https://raw.githubusercontent.com/user/repo/main/agent-handoffs.json"
-        )
+        print("  Example: https://raw.githubusercontent.com/user/repo/main/agent-handoffs.json")
         print()
         url = prompt_user("Template URL")
         if url:
@@ -2283,14 +2235,10 @@ def fetch_agent_template(url: str, template_type: str = "standard") -> Optional[
                 with open(template_path, "r") as f:
                     return f.read()
             except Exception as e:
-                print(
-                    f"{RED}❌ ERROR: Could not read {template_type} template: {e}{RESET}"
-                )
+                print(f"{RED}❌ ERROR: Could not read {template_type} template: {e}{RESET}")
                 return None
         else:
-            print(
-                f"{RED}❌ ERROR: {template_type} template not found in package{RESET}"
-            )
+            print(f"{RED}❌ ERROR: {template_type} template not found in package{RESET}")
             return None
 
     elif template_type == "custom" and url:
@@ -2390,9 +2338,11 @@ def agent_onboard(project_path: str = ".") -> int:
             return 0
 
     # 3. Interactive questions (4 max)
-    use_delegation = prompt_user(
-        "Use delegation/tasks/ structure? (recommended)", "Y"
-    ).lower() in ["y", "yes", ""]
+    use_delegation = prompt_user("Use delegation/tasks/ structure? (recommended)", "Y").lower() in [
+        "y",
+        "yes",
+        "",
+    ]
 
     organize_docs = prompt_user("Organize root docs into docs/?", "n").lower() in [
         "y",
@@ -2469,9 +2419,7 @@ def agent_onboard(project_path: str = ".") -> int:
                     print(
                         f"  {CYAN}ℹ️{RESET}  Original tasks/ preserved (remove manually if desired)"
                     )
-                    print(
-                        f"  {CYAN}ℹ️{RESET}  Rollback: rm -rf tasks && mv tasks.backup tasks"
-                    )
+                    print(f"  {CYAN}ℹ️{RESET}  Rollback: rm -rf tasks && mv tasks.backup tasks")
 
                 except Exception as e:
                     print(f"  {RED}❌{RESET} Migration failed: {e}")
@@ -2485,9 +2433,7 @@ def agent_onboard(project_path: str = ".") -> int:
         print(f"{BOLD}Documentation Organization:{RESET}")
 
         # Find markdown files in root
-        root_docs = [
-            f for f in os.listdir(".") if f.endswith(".md") and not f.startswith(".")
-        ]
+        root_docs = [f for f in os.listdir(".") if f.endswith(".md") and not f.startswith(".")]
 
         if len(root_docs) > 0:
             print(f"  Found {len(root_docs)} markdown file(s) in root")
@@ -2507,9 +2453,7 @@ def agent_onboard(project_path: str = ".") -> int:
                         moved_count += 1
 
                 if moved_count > 0:
-                    print(
-                        f"  {GREEN}✅{RESET} Organized {moved_count} doc(s) into docs/"
-                    )
+                    print(f"  {GREEN}✅{RESET} Organized {moved_count} doc(s) into docs/")
                 else:
                     print(f"  {CYAN}ℹ️{RESET}  No docs needed organizing")
 
@@ -2562,9 +2506,7 @@ def agent_onboard(project_path: str = ".") -> int:
                 print(f"  {RED}❌{RESET} Failed to fetch agent template")
                 return 1
         else:
-            print(
-                f"  {CYAN}ℹ️{RESET}  Skipped agent-handoffs.json (manual setup requested)"
-            )
+            print(f"  {CYAN}ℹ️{RESET}  Skipped agent-handoffs.json (manual setup requested)")
 
         # Render current-state.json
         current_state_template = templates_dir / "current-state.json.template"
@@ -2579,9 +2521,7 @@ def agent_onboard(project_path: str = ".") -> int:
         # Render README.md
         readme_template = templates_dir / "README.md.template"
         if readme_template.exists():
-            render_template(
-                str(readme_template), ".agent-context/README.md", template_vars
-            )
+            render_template(str(readme_template), ".agent-context/README.md", template_vars)
             print(f"  {GREEN}✅{RESET} Created .agent-context/README.md")
 
         # Copy AGENT-SYSTEM-GUIDE.md if it exists and isn't already there
@@ -2620,9 +2560,7 @@ def agent_onboard(project_path: str = ".") -> int:
 
         except Exception as e:
             print(f"  {YELLOW}⚠️{RESET}  Could not update config: {e}")
-            print(
-                f"     Manually set task_directory: delegation/tasks/ in .adversarial/config.yml"
-            )
+            print(f"     Manually set task_directory: delegation/tasks/ in .adversarial/config.yml")
 
     # 9. Update .gitignore
     print()
@@ -2676,9 +2614,7 @@ def agent_onboard(project_path: str = ".") -> int:
         verification_checks.append((f"current-state.json invalid: {e}", False))
 
     # Check directories exist
-    verification_checks.append(
-        (".agent-context/ exists", os.path.exists(".agent-context"))
-    )
+    verification_checks.append((".agent-context/ exists", os.path.exists(".agent-context")))
 
     if use_delegation:
         verification_checks.append(
@@ -2783,9 +2719,7 @@ def split(
 
         # Check if splitting is recommended
         if lines <= max_lines:
-            print(
-                f"{GREEN}✅ File is under recommended limit ({max_lines} lines){RESET}"
-            )
+            print(f"{GREEN}✅ File is under recommended limit ({max_lines} lines){RESET}")
             print("No splitting needed.")
             return 0
 
@@ -2803,9 +2737,7 @@ def split(
             splits = split_by_phases(content)
             print(f"\n💡 Suggested splits (by phases):")
         else:
-            print(
-                f"{RED}Error: Unknown strategy '{strategy}'. Use 'sections' or 'phases'.{RESET}"
-            )
+            print(f"{RED}Error: Unknown strategy '{strategy}'. Use 'sections' or 'phases'.{RESET}")
             return 1
 
         # Display split preview
@@ -2977,26 +2909,18 @@ For more information: https://github.com/movito/adversarial-workflow
     subparsers.add_parser("doctor", help="Alias for 'check'")
 
     # health command
-    health_parser = subparsers.add_parser(
-        "health", help="Comprehensive system health check"
-    )
+    health_parser = subparsers.add_parser("health", help="Comprehensive system health check")
     health_parser.add_argument(
         "--verbose", "-v", action="store_true", help="Show detailed diagnostics"
     )
-    health_parser.add_argument(
-        "--json", action="store_true", help="Output in JSON format"
-    )
+    health_parser.add_argument("--json", action="store_true", help="Output in JSON format")
 
     # agent command (with subcommands)
     agent_parser = subparsers.add_parser("agent", help="Agent coordination commands")
-    agent_subparsers = agent_parser.add_subparsers(
-        dest="agent_subcommand", help="Agent subcommand"
-    )
+    agent_subparsers = agent_parser.add_subparsers(dest="agent_subcommand", help="Agent subcommand")
 
     # agent onboard subcommand
-    onboard_parser = agent_subparsers.add_parser(
-        "onboard", help="Set up agent coordination system"
-    )
+    onboard_parser = agent_subparsers.add_parser("onboard", help="Set up agent coordination system")
     onboard_parser.add_argument(
         "--path", default=".", help="Project path (default: current directory)"
     )
@@ -3005,12 +2929,8 @@ For more information: https://github.com/movito/adversarial-workflow
     subparsers.add_parser("review", help="Run Phase 3: Code review")
 
     # validate command
-    validate_parser = subparsers.add_parser(
-        "validate", help="Run Phase 4: Test validation"
-    )
-    validate_parser.add_argument(
-        "test_command", nargs="?", help="Test command to run (optional)"
-    )
+    validate_parser = subparsers.add_parser("validate", help="Run Phase 4: Test validation")
+    validate_parser.add_argument("test_command", nargs="?", help="Test command to run (optional)")
 
     # split command
     split_parser = subparsers.add_parser(
@@ -3056,9 +2976,7 @@ For more information: https://github.com/movito/adversarial-workflow
             # Only warn for user-defined evaluators, not built-ins
             # Built-in conflicts are intentional (e.g., 'review' command vs 'review' evaluator)
             if getattr(config, "source", None) != "builtin":
-                logger.warning(
-                    "Evaluator '%s' conflicts with CLI command; skipping", name
-                )
+                logger.warning("Evaluator '%s' conflicts with CLI command; skipping", name)
             # Mark as registered to prevent alias re-registration attempts
             registered_configs.add(id(config))
             continue

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/adversarial_workflow/evaluators/discovery.py

@@ -51,9 +51,7 @@ def parse_evaluator_yaml(yml_file: Path) -> EvaluatorConfig:
 
     # Ensure parsed data is a dict (YAML can parse scalars, lists, etc.)
     if not isinstance(data, dict):
-        raise EvaluatorParseError(
-            f"YAML must be a mapping, got {type(data).__name__}: {yml_file}"
-        )
+        raise EvaluatorParseError(f"YAML must be a mapping, got {type(data).__name__}: {yml_file}")
 
     # Validate required fields exist
     required = [
@@ -91,9 +89,7 @@ def parse_evaluator_yaml(yml_file: Path) -> EvaluatorConfig:
     elif isinstance(aliases, str):
         data["aliases"] = [aliases]
     elif not isinstance(aliases, list):
-        raise EvaluatorParseError(
-            f"aliases must be string or list, got {type(aliases).__name__}"
-        )
+        raise EvaluatorParseError(f"aliases must be string or list, got {type(aliases).__name__}")
 
     # Validate alias names - must be strings with valid format
     for alias in data.get("aliases", []):
@@ -131,18 +127,14 @@ def parse_evaluator_yaml(yml_file: Path) -> EvaluatorConfig:
         # Check for bool before int (bool is subclass of int in Python)
         # YAML parses 'yes'/'true' as True, 'no'/'false' as False
         if isinstance(timeout, bool):
-            raise EvaluatorParseError(
-                f"Field 'timeout' must be an integer, got bool: {timeout!r}"
-            )
+            raise EvaluatorParseError(f"Field 'timeout' must be an integer, got bool: {timeout!r}")
         if not isinstance(timeout, int):
             raise EvaluatorParseError(
                 f"Field 'timeout' must be an integer, got {type(timeout).__name__}: {timeout!r}"
             )
         # timeout=0 is invalid (does not disable timeout - use a large value instead)
         if timeout <= 0:
-            raise EvaluatorParseError(
-                f"Field 'timeout' must be positive (> 0), got {timeout}"
-            )
+            raise EvaluatorParseError(f"Field 'timeout' must be positive (> 0), got {timeout}")
         if timeout > 600:
             logger.warning(
                 "Timeout %ds exceeds maximum (600s), clamping to 600s in %s",
@@ -167,9 +159,7 @@ def parse_evaluator_yaml(yml_file: Path) -> EvaluatorConfig:
     }
     unknown = set(data.keys()) - known_fields
     if unknown:
-        logger.warning(
-            "Unknown fields in %s: %s", yml_file.name, ", ".join(sorted(unknown))
-        )
+        logger.warning("Unknown fields in %s: %s", yml_file.name, ", ".join(sorted(unknown)))
 
     # Build filtered data dict
     filtered_data = {k: v for k, v in data.items() if k in known_fields}

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/adversarial_workflow/evaluators/runner.py

@@ -227,10 +227,7 @@ def _execute_script(
 
     # Validate output
     file_basename = Path(file_path).stem
-    log_file = (
-        Path(project_config["log_directory"])
-        / f"{file_basename}-{config.output_suffix}.md"
-    )
+    log_file = Path(project_config["log_directory"]) / f"{file_basename}-{config.output_suffix}.md"
 
     is_valid, verdict, message = validate_evaluation_output(str(log_file))
 
@@ -241,9 +238,7 @@ def _execute_script(
     return _report_verdict(verdict, log_file, config)
 
 
-def _report_verdict(
-    verdict: str | None, log_file: Path, config: EvaluatorConfig
-) -> int:
+def _report_verdict(verdict: str | None, log_file: Path, config: EvaluatorConfig) -> int:
     """Report the evaluation verdict to terminal."""
     print()
     if verdict == "APPROVED":

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/adversarial_workflow/utils/config.py

@@ -24,9 +24,7 @@ def load_config(config_path: str = ".adversarial/config.yml") -> dict[str, Any]:
         with open(config_path) as f:
             file_config = yaml.safe_load(f) or {}
             if not isinstance(file_config, dict):
-                raise ValueError(
-                    f"Config file must be a mapping, got {type(file_config).__name__}"
-                )
+                raise ValueError(f"Config file must be a mapping, got {type(file_config).__name__}")
             config.update(file_config)
 
     # Override with environment variables

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/adversarial_workflow/utils/file_splitter.py

@@ -141,9 +141,7 @@ def split_by_sections(content: str, max_lines: int = 500) -> List[Dict[str, Any]
         is_section_boundary = re.match(r"^#+\s+", line.strip())
         approaching_limit = len(current_split_lines) >= max_lines * 0.8
 
-        if len(current_split_lines) >= max_lines or (
-            is_section_boundary and approaching_limit
-        ):
+        if len(current_split_lines) >= max_lines or (is_section_boundary and approaching_limit):
             # Create split
             split_content = "\n".join(current_split_lines)
             splits.append(
@@ -318,9 +316,7 @@ def split_at_lines(content: str, line_numbers: List[int]) -> List[Dict[str, Any]
     return splits
 
 
-def generate_split_files(
-    original: str, splits: List[Dict[str, Any]], output_dir: str
-) -> List[str]:
+def generate_split_files(original: str, splits: List[Dict[str, Any]], output_dir: str) -> List[str]:
     """Generate split files with metadata and cross-references.
 
     Args:

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/adversarial_workflow/utils/validation.py

@@ -47,9 +47,7 @@ def validate_evaluation_output(
         "concerns",
     ]
 
-    has_evaluation_content = any(
-        marker in content_lower for marker in evaluation_markers
-    )
+    has_evaluation_content = any(marker in content_lower for marker in evaluation_markers)
     if not has_evaluation_content:
         return (
             False,

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/adversarial_workflow.egg-info/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: adversarial-workflow
-Version: 0.6.3
-Summary: Multi-stage AI code review system preventing phantom work - Author/Evaluator pattern
+Version: 0.6.4
+Summary: Multi-stage AI evaluation system for task plans, code review, and test validation
 Author: Fredrik Matheson
 License: MIT
 Project-URL: Homepage, https://github.com/movito/adversarial-workflow

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/pyproject.toml

@@ -5,9 +5,9 @@ build-backend = "setuptools.build_meta"
 [project]
 name = "adversarial-workflow"
 
-version = "0.6.3"
+version = "0.6.4"
 
-description = "Multi-stage AI code review system preventing phantom work - Author/Evaluator pattern"
+description = "Multi-stage AI evaluation system for task plans, code review, and test validation"
 readme = "README.md"
 authors = [
     {name = "Fredrik Matheson"}

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/tests/test_cli_dynamic_commands.py

@@ -78,9 +78,7 @@ output_suffix: CUSTOM-TEST
         monkeypatch.chdir(tmp_path)
 
         result = run_cli(["--help"], cwd=tmp_path)
-        assert (
-            "custom" in result.stdout
-        ), f"'custom' not found in help output:\n{result.stdout}"
+        assert "custom" in result.stdout, f"'custom' not found in help output:\n{result.stdout}"
         assert "Custom test evaluator" in result.stdout
 
     def test_multiple_local_evaluators_in_help(self, tmp_path, monkeypatch, run_cli):
@@ -351,9 +349,7 @@ class TestBackwardsCompatibility:
 class TestGracefulDegradation:
     """Test graceful degradation on errors."""
 
-    def test_help_works_without_local_evaluators_dir(
-        self, tmp_path, monkeypatch, run_cli
-    ):
+    def test_help_works_without_local_evaluators_dir(self, tmp_path, monkeypatch, run_cli):
         """CLI help works even without .adversarial/evaluators/ directory."""
         adv_dir = tmp_path / ".adversarial"
         adv_dir.mkdir(parents=True)
@@ -422,9 +418,7 @@ class TestReviewCommandBackwardsCompatibility:
         # Review should NOT have --timeout flag (that's for evaluators)
         assert "--timeout" not in result.stdout
 
-    def test_review_command_not_overridden_by_evaluator(
-        self, tmp_path, monkeypatch, run_cli
-    ):
+    def test_review_command_not_overridden_by_evaluator(self, tmp_path, monkeypatch, run_cli):
         """Review command cannot be overridden by local evaluator."""
         adv_dir = tmp_path / ".adversarial"
         adv_dir.mkdir(parents=True)
@@ -492,9 +486,7 @@ aliases:
         assert "--path" in result_init.stdout
         assert "--interactive" in result_init.stdout
 
-    def test_evaluator_with_conflicting_name_and_alias(
-        self, tmp_path, monkeypatch, run_cli
-    ):
+    def test_evaluator_with_conflicting_name_and_alias(self, tmp_path, monkeypatch, run_cli):
         """Evaluator with conflicting name doesn't crash when alias is processed."""
         adv_dir = tmp_path / ".adversarial"
         adv_dir.mkdir(parents=True)

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/tests/test_env_loading.py

@@ -94,9 +94,7 @@ class TestCheckEnvCount:
         """
         # Create .env with 3 variables
         (tmp_path / ".env").write_text(
-            "OPENAI_API_KEY=sk-test\n"
-            "ANTHROPIC_API_KEY=ant-test\n"
-            "CUSTOM_KEY=custom-value\n"
+            "OPENAI_API_KEY=sk-test\n" "ANTHROPIC_API_KEY=ant-test\n" "CUSTOM_KEY=custom-value\n"
         )
 
         # Remove keys from environment to isolate test

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/tests/test_evaluate.py

@@ -49,9 +49,7 @@ class TestEvaluate:
 
     @patch("shutil.which")
     @patch("adversarial_workflow.cli.load_config")
-    def test_evaluate_aider_not_found(
-        self, mock_load_config, mock_which, tmp_path, capsys
-    ):
+    def test_evaluate_aider_not_found(self, mock_load_config, mock_which, tmp_path, capsys):
         """Test evaluate when aider is not available."""
         # Create a test file
         task_file = tmp_path / "test_task.md"
@@ -411,9 +409,7 @@ class TestVerifyTokenCount:
 
     @patch("adversarial_workflow.cli.estimate_file_tokens")
     @patch("adversarial_workflow.cli.extract_token_count_from_log")
-    def test_verify_token_count_normal(
-        self, mock_extract, mock_estimate, tmp_path, capsys
-    ):
+    def test_verify_token_count_normal(self, mock_extract, mock_estimate, tmp_path, capsys):
         """Test normal token count verification."""
         task_file = tmp_path / "task.md"
         log_file = tmp_path / "log.md"
@@ -433,9 +429,7 @@ class TestVerifyTokenCount:
 
     @patch("adversarial_workflow.cli.estimate_file_tokens")
     @patch("adversarial_workflow.cli.extract_token_count_from_log")
-    def test_verify_token_count_low_warning(
-        self, mock_extract, mock_estimate, tmp_path, capsys
-    ):
+    def test_verify_token_count_low_warning(self, mock_extract, mock_estimate, tmp_path, capsys):
         """Test token count verification warns on suspiciously low usage."""
         task_file = tmp_path / "task.md"
         log_file = tmp_path / "log.md"

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/tests/test_evaluator_discovery.py

@@ -388,9 +388,7 @@ fallback_model: yes
 """
         )
 
-        with pytest.raises(
-            EvaluatorParseError, match="'fallback_model' must be a string"
-        ):
+        with pytest.raises(EvaluatorParseError, match="'fallback_model' must be a string"):
             parse_evaluator_yaml(yml)
 
     def test_parse_with_valid_timeout(self, tmp_path):

{adversarial_workflow-0.6.3 → adversarial_workflow-0.6.4}/tests/test_list_evaluators.py

@@ -53,9 +53,7 @@ aliases:
         result = run_cli(["--help"])
         assert "list-evaluators" in result.stdout
 
-    def test_list_evaluators_skips_alias_duplicates(
-        self, tmp_path, monkeypatch, run_cli
-    ):
+    def test_list_evaluators_skips_alias_duplicates(self, tmp_path, monkeypatch, run_cli):
         """Aliases do not cause duplicate entries in output."""
         eval_dir = tmp_path / ".adversarial" / "evaluators"
         eval_dir.mkdir(parents=True)
@@ -80,9 +78,7 @@ aliases:
         assert result.stdout.count("Knowledge evaluation") == 1
         assert "aliases: knowledge, research" in result.stdout
 
-    def test_list_evaluators_shows_version_if_not_default(
-        self, tmp_path, monkeypatch, run_cli
-    ):
+    def test_list_evaluators_shows_version_if_not_default(self, tmp_path, monkeypatch, run_cli):
         """Shows version only when it differs from default 1.0.0."""
         eval_dir = tmp_path / ".adversarial" / "evaluators"
         eval_dir.mkdir(parents=True)