fix: Windows command line length limit errors

.claude/templates/coding_prompt.template.md

@@ -371,19 +371,22 @@ feature_get_stats
 # 2. Get the NEXT feature to work on (one feature only)
 feature_get_next
 
-# 3. Mark a feature as in-progress (call immediately after feature_get_next)
+# 3. Get a specific feature by ID (for parallel mode where feature is pre-assigned)
+feature_get_by_id with feature_id={id}
+
+# 4. Mark a feature as in-progress (call immediately after feature_get_next)
 feature_mark_in_progress with feature_id={id}
 
-# 4. Get up to 3 random passing features for regression testing
+# 5. Get up to 3 random passing features for regression testing
 feature_get_for_regression
 
-# 5. Mark a feature as passing (after verification)
+# 6. Mark a feature as passing (after verification)
 feature_mark_passing with feature_id={id}
 
-# 6. Skip a feature (moves to end of queue) - ONLY when blocked by dependency
+# 7. Skip a feature (moves to end of queue) - ONLY when blocked by dependency
 feature_skip with feature_id={id}
 
-# 7. Clear in-progress status (when abandoning a feature)
+# 8. Clear in-progress status (when abandoning a feature)
 feature_clear_in_progress with feature_id={id}
 ```
 

.claude/templates/coding_prompt_yolo.template.md

@@ -205,16 +205,19 @@ feature_get_stats
 # 2. Get the NEXT feature to work on (one feature only)
 feature_get_next
 
-# 3. Mark a feature as in-progress (call immediately after feature_get_next)
+# 3. Get a specific feature by ID (for parallel mode where feature is pre-assigned)
+feature_get_by_id with feature_id={id}
+
+# 4. Mark a feature as in-progress (call immediately after feature_get_next)
 feature_mark_in_progress with feature_id={id}
 
-# 4. Mark a feature as passing (after lint/type-check succeeds)
+# 5. Mark a feature as passing (after lint/type-check succeeds)
 feature_mark_passing with feature_id={id}
 
-# 5. Skip a feature (moves to end of queue) - ONLY when blocked by dependency
+# 6. Skip a feature (moves to end of queue) - ONLY when blocked by dependency
 feature_skip with feature_id={id}
 
-# 6. Clear in-progress status (when abandoning a feature)
+# 7. Clear in-progress status (when abandoning a feature)
 feature_clear_in_progress with feature_id={id}
 ```
 

agent.py

@@ -25,8 +25,8 @@
  get_initializer_prompt,
  get_coding_prompt,
  get_coding_prompt_yolo,
+ get_coding_prompt_parallel,
  copy_spec_to_project,
- has_project_prompts,
 )
 
 
@@ -113,20 +113,35 @@ async def run_autonomous_agent(
  model: str,
  max_iterations: Optional[int] = None,
  yolo_mode: bool = False,
+ work_dir: Optional[Path] = None,
+ feature_id: Optional[int] = None,
 ) -> None:
  """
  Run the autonomous agent loop.
 
  Args:
- project_dir: Directory for the project
+ project_dir: Directory for prompts and features.db
  model: Claude model to use
  max_iterations: Maximum number of iterations (None for unlimited)
  yolo_mode: If True, skip browser testing and use YOLO prompt
+ work_dir: Directory for code changes (default: same as project_dir)
+ In parallel mode, this is the git worktree path.
+ feature_id: If provided, work on this specific feature (parallel mode).
+ Worker is bound to this feature and won't use feature_get_next.
  """
+ # work_dir defaults to project_dir if not specified
+ if work_dir is None:
+ work_dir = project_dir
+
+ is_parallel_mode = feature_id is not None
+
  print("\n" + "=" * 70)
  print(" AUTONOMOUS CODING AGENT DEMO")
  print("=" * 70)
  print(f"\nProject directory: {project_dir}")
+ if is_parallel_mode:
+ print(f"Worktree path: {work_dir}")
+ print(f"Bound to feature ID: {feature_id}")
  print(f"Model: {model}")
  if yolo_mode:
  print("Mode: YOLO (testing disabled)")
@@ -138,8 +153,9 @@ async def run_autonomous_agent(
  print("Max iterations: Unlimited (will run until completion)")
  print()
 
- # Create project directory
+ # Create directories
  project_dir.mkdir(parents=True, exist_ok=True)
+ work_dir.mkdir(parents=True, exist_ok=True)
 
  # Check if this is a fresh start or continuation
  # Uses has_features() which checks if the database actually has features,
@@ -177,15 +193,19 @@ async def run_autonomous_agent(
  print_session_header(iteration, is_first_run)
 
  # Create client (fresh context)
- client = create_client(project_dir, model, yolo_mode=yolo_mode)
+ # In parallel mode, work_dir is the worktree; project_dir has DB/prompts
+ client = create_client(project_dir, model, yolo_mode=yolo_mode, work_dir=work_dir)
 
  # Choose prompt based on session type
  # Pass project_dir to enable project-specific prompts
  if is_first_run:
  prompt = get_initializer_prompt(project_dir)
  is_first_run = False # Only use initializer once
+ elif is_parallel_mode:
+ # Parallel mode: use feature_id-bound prompt
+ prompt = get_coding_prompt_parallel(project_dir, feature_id, yolo_mode)
  else:
- # Use YOLO prompt if in YOLO mode
+ # Single agent mode: use YOLO or standard prompt
  if yolo_mode:
  prompt = get_coding_prompt_yolo(project_dir)
  else:

api/database.py

@@ -5,17 +5,28 @@
 SQLite database schema for feature storage using SQLAlchemy.
 """
 
+import enum
 from pathlib import Path
 from typing import Optional
 
-from sqlalchemy import Boolean, Column, Integer, String, Text, create_engine
+from sqlalchemy import Boolean, Column, DateTime, Enum, Integer, String, Text, create_engine
 from sqlalchemy.ext.declarative import declarative_base
 from sqlalchemy.orm import sessionmaker, Session
 from sqlalchemy.types import JSON
 
 Base = declarative_base()
 
 
+class FeatureStatus(enum.Enum):
+ """Status of a feature in the queue."""
+
+ PENDING = "pending" # Available for claiming
+ IN_PROGRESS = "in_progress" # Claimed by a worker (leased)
+ PASSING = "passing" # Completed successfully
+ CONFLICT = "conflict" # Merge conflict; manual resolution needed
+ FAILED = "failed" # Agent could not complete (permanent)
+
+
 class Feature(Base):
  """Feature model representing a test case/feature to implement."""
 
@@ -27,9 +38,26 @@ class Feature(Base):
  name = Column(String(255), nullable=False)
  description = Column(Text, nullable=False)
  steps = Column(JSON, nullable=False) # Stored as JSON array
+
+ # Status tracking (new enum-based status for parallel execution)
+ status = Column(
+ Enum(FeatureStatus, values_callable=lambda x: [e.value for e in x]),
+ default=FeatureStatus.PENDING,
+ index=True,
+ )
+
+ # Legacy fields - kept for backward compatibility
  passes = Column(Boolean, default=False, index=True)
  in_progress = Column(Boolean, default=False, index=True)
 
+ # Claim/lease tracking for parallel execution
+ claimed_by = Column(String(100), nullable=True) # Worker ID holding this feature
+ claimed_at = Column(DateTime, nullable=True) # Timestamp for lease expiry detection
+
+ # Completion audit
+ completed_at = Column(DateTime, nullable=True)
+ completed_by = Column(String(100), nullable=True)
+
  def to_dict(self) -> dict:
  """Convert feature to dictionary for JSON serialization."""
  return {
@@ -41,6 +69,11 @@ def to_dict(self) -> dict:
  "steps": self.steps,
  "passes": self.passes,
  "in_progress": self.in_progress,
+ "status": self.status.value if self.status else FeatureStatus.PENDING.value,
+ "claimed_by": self.claimed_by,
+ "claimed_at": self.claimed_at.isoformat() if self.claimed_at else None,
+ "completed_at": self.completed_at.isoformat() if self.completed_at else None,
+ "completed_by": self.completed_by,
  }
 
 
@@ -58,19 +91,46 @@ def get_database_url(project_dir: Path) -> str:
  return f"sqlite:///{db_path.as_posix()}"
 
 
-def _migrate_add_in_progress_column(engine) -> None:
- """Add in_progress column to existing databases that don't have it."""
+def _migrate_database_schema(engine) -> None:
+ """Migrate existing databases to add new columns for parallel execution."""
  from sqlalchemy import text
 
  with engine.connect() as conn:
- # Check if column exists
+ # Check existing columns
  result = conn.execute(text("PRAGMA table_info(features)"))
  columns = [row[1] for row in result.fetchall()]
 
+ # Migration: in_progress column (legacy)
  if "in_progress" not in columns:
- # Add the column with default value
  conn.execute(text("ALTER TABLE features ADD COLUMN in_progress BOOLEAN DEFAULT 0"))
- conn.commit()
+
+ # Migration: status column for parallel execution
+ if "status" not in columns:
+ conn.execute(text("ALTER TABLE features ADD COLUMN status TEXT DEFAULT 'pending'"))
+ # Migrate existing data: passes=True -> 'passing', in_progress=True -> 'in_progress'
+ conn.execute(text("""
+ UPDATE features SET status = CASE
+ WHEN passes = 1 THEN 'passing'
+ WHEN in_progress = 1 THEN 'in_progress'
+ ELSE 'pending'
+ END
+ """))
+
+ # Migration: claim tracking columns
+ if "claimed_by" not in columns:
+ conn.execute(text("ALTER TABLE features ADD COLUMN claimed_by TEXT"))
+
+ if "claimed_at" not in columns:
+ conn.execute(text("ALTER TABLE features ADD COLUMN claimed_at DATETIME"))
+
+ # Migration: completion audit columns
+ if "completed_at" not in columns:
+ conn.execute(text("ALTER TABLE features ADD COLUMN completed_at DATETIME"))
+
+ if "completed_by" not in columns:
+ conn.execute(text("ALTER TABLE features ADD COLUMN completed_by TEXT"))
+
+ conn.commit()
 
 
 def create_database(project_dir: Path) -> tuple:
@@ -87,8 +147,8 @@ def create_database(project_dir: Path) -> tuple:
  engine = create_engine(db_url, connect_args={"check_same_thread": False})
  Base.metadata.create_all(bind=engine)
 
- # Migrate existing databases to add in_progress column
-_migrate_add_in_progress_column(engine)
+ # Migrate existing databases to add new columns
+_migrate_database_schema(engine)
 
  SessionLocal = sessionmaker(autocommit=False, autoflush=False, bind=engine)
  return engine, SessionLocal

autonomous_agent_demo.py

@@ -23,7 +23,6 @@
 
 import argparse
 import asyncio
-import os
 from pathlib import Path
 
 from dotenv import load_dotenv
@@ -97,6 +96,21 @@ def parse_args() -> argparse.Namespace:
  help="Enable YOLO mode: rapid prototyping without browser testing",
  )
 
+ # Parallel execution arguments (used by parallel_coordinator.py)
+ parser.add_argument(
+ "--worktree-path",
+ type=str,
+ default=None,
+ help="Git worktree path for code changes (parallel mode only)",
+ )
+
+ parser.add_argument(
+ "--feature-id",
+ type=int,
+ default=None,
+ help="Feature ID to implement (parallel mode only - binds to claimed feature)",
+ )
+
  return parser.parse_args()
 
 
@@ -128,14 +142,19 @@ def main() -> None:
  print("Use an absolute path or register the project first.")
  return
 
+ # In parallel mode, use worktree for code changes
+ work_dir = Path(args.worktree_path) if args.worktree_path else project_dir
+
  try:
  # Run the agent (MCP server handles feature database)
  asyncio.run(
  run_autonomous_agent(
- project_dir=project_dir,
+ project_dir=project_dir, # For DB/prompts
+ work_dir=work_dir, # For code changes (may be worktree)
  model=args.model,
  max_iterations=args.max_iterations,
  yolo_mode=args.yolo,
+ feature_id=args.feature_id, # Bound to claimed feature (parallel mode)
  )
  )
  except KeyboardInterrupt: