feat: multi-process prefork worker pool

[pratyush618]

Summary

• Prefork worker pool: queue.run_worker(pool="prefork", app="myapp:queue") spawns child Python processes with independent GILs for true CPU parallelism
• Rust PreforkPool: implements WorkerDispatcher trait, dispatches serialized jobs over stdin pipes, reads results from stdout pipes
• JSON Lines IPC: ParentMessage (job/shutdown) and ChildMessage (success/failure/cancelled/ready) over stdio
• Child process: python -m taskito.prefork <app_path> — imports app, builds task registry, executes tasks in a read-execute-write loop
• Least-loaded dispatch: jobs sent to the child with fewest in-flight tasks
• Graceful shutdown: sends shutdown message to children, waits with timeout, kills if needed

Architecture

Parent: Scheduler → PreforkPool → stdin pipes → Children
Children: stdin → execute task → stdout → Reader threads → result_tx → Parent

Test plan

• cargo check all feature combos
• cargo test --workspace — 43 Rust tests pass
• uv run python -m pytest tests/python/ -v — 352 pass, 9 skipped
• Prefork child process starts and responds to jobs manually
• pool="prefork" validates app= is required
• pool="thread" (default) unchanged