Add closed-loop atomic-action execution runner

[yuecideng]

Description

This PR adds a closed-loop execution layer for typed atomic-action plans. It introduces a non-blocking ExecutionRunner, explicit observation/command/clock ports, command acknowledgement and cancellation semantics, trajectory timing propagation, safe hold behavior, and a simulation adapter that preserves batched environment correlation.

The runner closes the gap between the planning/recovery contracts introduced by #448 and real simulation execution. It also adds a tracking-error recovery tutorial that demonstrates observation-driven invalidation, replanning, and resumed execution in a headless simulation.

Dependencies: #448. This is a stacked PR whose base is refactor/atomic-action-pr1-contracts; review the single PR2A commit rather than the full PR1 diff.

Fixes: N/A (no linked issue).

Type of change

• New feature (non-breaking change which adds functionality)

Validation

• black --check --diff --color ./
• Targeted atomic-action tests: 164 passed, 2 skipped, 5 deselected
• Headless simulation smoke tests with one and two environments; an injected 0.35 rad tracking error emitted tracking_error, replanned, and completed after 83 accepted commands
• Sphinx HTML build completed successfully; the build retained 618 pre-existing repository-wide warnings
• Python compile and public export checks passed

Screenshots

Not applicable; this change adds runtime APIs and a headless simulation tutorial.

Checklist

• I have run the black . command to format the code base.
• I have made corresponding changes to the documentation
• I have added tests that prove my fix is effective or that my feature works
• Dependencies have been updated, if applicable. (No dependency-file changes required.)