claw/docs/superpowers/plans/2026-04-20-scene-skill-102-full-direct-mock-execution-plan.md

Scene Skill 102 Full Direct Mock Execution Plan

Date: 2026-04-20 Status: Draft Upstream Design: docs/superpowers/specs/2026-04-20-scene-skill-102-full-direct-mock-execution-design.md

Plan Intent

Run all 102 final materialized scene skill scripts through a local direct mock runtime.

This plan expands beyond representative harness execution, but remains fully mock-only and local.

Fixed Inputs

1. examples/scene_skill_102_final_materialization_2026-04-19/skills
2. examples/scene_skill_102_final_materialization_2026-04-19/scene_skill_102_index.json
3. tests/fixtures/generated_scene/scene_skill_102_static_validation_2026-04-20.json
4. tests/fixtures/generated_scene/scene_skill_102_mock_runtime_harness_results_2026-04-20.json

Planned Outputs

1. tests/fixtures/generated_scene/scene_skill_102_full_direct_mock_execution_2026-04-20.json
2. docs/superpowers/reports/2026-04-20-scene-skill-102-full-direct-mock-execution-report.md

Allowed Files

1. new direct mock runner under tests/
2. tests/fixtures/generated_scene/scene_skill_102_full_direct_mock_execution_2026-04-20.json
3. docs/superpowers/reports/2026-04-20-scene-skill-102-full-direct-mock-execution-report.md

Forbidden Files

1. src/generated_scene/analyzer.rs
2. src/generated_scene/generator.rs
3. src/generated_scene/ir.rs
4. examples/scene_skill_102_final_materialization_2026-04-19/skills/**
5. tests/fixtures/generated_scene/scene_execution_board_2026-04-18.json

Phase 0: Freeze Direct Mock Boundary

Tasks

1. Confirm representative mock harness is complete.
2. Confirm this plan does not mutate generated skill packages.
3. Confirm this plan does not use real network, browser, or credentials.

Acceptance Criteria

1. direct mock starts from final materialized skills
2. generated skills remain unchanged

Phase 1: Build Direct Mock Runner

Tasks

1. load the 102 scene index
2. locate each generated script
3. reuse fake runtime dependencies by archetype
4. call buildBrowserEntrypointResult
5. capture artifact status, row count, failure reason, and mock request log

Acceptance Criteria

1. every scene is attempted
2. no single scene failure aborts the full run
3. no real request is sent

Phase 2: Execute Direct Mock For 102

Tasks

1. run the direct mock runner
2. write per-scene direct mock result
3. classify each scene as:
   • direct-mock-pass
   • direct-mock-partial
   • direct-mock-fail

Acceptance Criteria

1. output record count is 102
2. each failure has a named reason

Phase 3: Publish Report

Tasks

1. summarize direct mock pass/fail
2. summarize results by archetype
3. identify remaining mock-only blockers
4. recommend whether pseudo-production batch selection should start

Acceptance Criteria

1. report does not claim production execution
2. report separates mock pass from production pass

Completion Criteria

This plan is complete when:

1. all 102 scenes have direct mock results
2. JSON asset is published
3. report is published
4. generated skill packages remain unchanged

Stop Statement

Stop after publishing direct mock execution results and report.

Do not start pseudo-production batch selection under this plan.