# Scene Skill 102 Parameter Dictionary And Invocation Template Normalization Plan

> Date: 2026-04-20
> Design: `2026-04-20-scene-skill-102-parameter-dictionary-template-normalization-design.md`

## Plan Intent

Close the natural-language parameter readiness gap for the fixed `10` required-param scene skills and refresh pseudo-production invocation templates.

## Fixed Input Bucket

The only input bucket is the `10` scenes marked `parameter-gap` in:

`tests/fixtures/generated_scene/scene_skill_102_natural_language_parameter_readiness_2026-04-20.json`

## Allowed Files

1. `examples/scene_skill_102_final_materialization_2026-04-19/skills/{fixed-10}/references/org-dictionary.json`
2. `tests/fixtures/generated_scene/scene_skill_102_natural_language_parameter_readiness_2026-04-20.json`
3. `tests/fixtures/generated_scene/scene_skill_102_natural_language_invocation_samples_2026-04-20.json`
4. `tests/fixtures/generated_scene/scene_skill_102_pseudoprod_execution_handoff_2026-04-20.json`
5. `tests/fixtures/generated_scene/scene_skill_102_parameter_dictionary_template_normalization_2026-04-20.json`
6. `docs/superpowers/reports/2026-04-20-scene-skill-102-parameter-dictionary-template-normalization-report.md`

## Forbidden Files

1. `src/compat/scene_platform/dispatch.rs`
2. `src/compat/scene_platform/resolvers.rs`
3. `src/generated_scene/analyzer.rs`
4. `src/generated_scene/generator.rs`
5. `examples/scene_skill_102_final_materialization_2026-04-19/skills/*/scripts/*`
6. `tests/fixtures/generated_scene/scene_execution_board_2026-04-18.json`

## Phase 0: Freeze Scope

### Tasks

1. Confirm the fixed `10` required-param scenes.
2. Confirm all current gaps are empty org dictionaries.
3. Confirm no runtime code changes are needed.

### Acceptance Criteria

1. No non-param scene is changed.
2. No browser or production execution is started.

## Phase 1: Populate Starter Organization Dictionaries

### Tasks

1. Write the pseudo-production starter dictionary into each fixed `10` skill.
2. Use the already-tested aliases:
   - `兰州公司`
   - `兰州供电公司`
   - `国网兰州供电公司`
   - `城关供电分公司`
   - `城关分公司`
   - `天水公司`
   - `天水供电公司`
   - `国网天水供电公司`
3. Mark dictionary provenance as starter, not full production.

### Acceptance Criteria

1. All fixed `10` dictionaries are non-empty arrays.
2. Each dictionary contains alias coverage for `兰州公司`.

## Phase 2: Refresh Parameter Readiness

### Tasks

1. Re-scan all `102` skills.
2. Recompute parameter readiness.
3. Verify the fixed `10` move to `parameter-ready`.
4. Keep implicit-risk classification for no-param scenes.

### Acceptance Criteria

1. `parameter-gap = 0`.
2. `parameter-ready = 10`.
3. `total scenes = 102`.

## Phase 3: Refresh Invocation Templates

### Tasks

1. Generate parameterized samples for the fixed `10`.
2. Ensure samples include concrete period, e.g. `月累计 2026-03`.
3. Ensure samples keep `。。。` suffix.

### Acceptance Criteria

1. All fixed `10` have parameterized sample input.
2. No-param scenes keep minimal invocation samples.

## Phase 4: Refresh Pseudo-Production Handoff

### Tasks

1. Update the selected pseudo-production handoff entries that are in the fixed `10`.
2. Replace bare scene-name inputs with parameterized inputs.
3. Preserve credential policy and evidence collection fields.

### Acceptance Criteria

1. Selected required-param scenes no longer use bare `场景名。。。` in handoff.
2. No credentials are written to the repository.

## Phase 5: Publish Normalization Report

### Tasks

1. Publish normalization JSON.
2. Publish superpowers report.
3. State remaining limits explicitly.

### Acceptance Criteria

1. Report states that dictionaries are starter dictionaries, not complete production unit trees.
2. Report states next step for pseudo-production execution preparation refresh.

## Stop Statement

Stop after dictionaries, readiness assets, invocation samples, handoff, and report are refreshed. Do not run browser, production, or runtime implementation work under this plan.