feat(rpgkit): add consolidated /rpgkit.plan workflow

RPG-Kit/README.hi-IN.md

@@ -137,11 +137,7 @@ uvx --from "git+https://github.com/microsoft/RPG-ZeroRepo.git#subdirectory=RPG-K
    /rpgkit.feature_build
    /rpgkit.feature_refactor
    [Optional] /rpgkit.feature_edit <edit instructions>
-   /rpgkit.build_skeleton
-   /rpgkit.build_data_flow
-   /rpgkit.design_base_classes
-   /rpgkit.design_interfaces
-   /rpgkit.plan_tasks
+   /rpgkit.plan
    /rpgkit.code_gen
    [Optional] /rpgkit.rpg_edit <edit instructions>
    ```
@@ -241,7 +237,7 @@ rpgkit update
 
 ## आगामी सुविधाएँ
 
-- **सरल जनरेशन कमांड्स:** वर्तमान बहु-चरण जनरेशन प्रवाह को कम कमांड्स में मर्ज किया जाएगा, जैसे `/rpgkit.generate_repo`, `/rpgkit.generate_feature` और `/rpgkit.plan`।
+- **सरल जनरेशन कमांड्स:** वर्तमान बहु-चरण जनरेशन प्रवाह को कम कमांड्स में मर्ज किया जाएगा, जैसे `/rpgkit.generate_repo` और `/rpgkit.generate_feature`। `/rpgkit.plan` 0.1.4 में रिलीज़ हो चुका है।
 - **बहु-भाषा समर्थन:** Go, C++, Rust, JavaScript/TypeScript और अन्य के लिए समर्थन जोड़ा जाएगा।
 - **अधिक प्लेटफ़ॉर्म एकीकरण:** विभिन्न सिस्टम्स पर विभिन्न AI कोडिंग एजेंट्स के लिए CLI और VS Code एक्सटेंशन वर्कफ़्लो में RPG-Kit समर्थन।
 

RPG-Kit/README.ja-JP.md

@@ -137,11 +137,7 @@ uvx --from "git+https://github.com/microsoft/RPG-ZeroRepo.git#subdirectory=RPG-K
    /rpgkit.feature_build
    /rpgkit.feature_refactor
    [Optional] /rpgkit.feature_edit <edit instructions>
-   /rpgkit.build_skeleton
-   /rpgkit.build_data_flow
-   /rpgkit.design_base_classes
-   /rpgkit.design_interfaces
-   /rpgkit.plan_tasks
+   /rpgkit.plan
    /rpgkit.code_gen
    [Optional] /rpgkit.rpg_edit <edit instructions>
    ```
@@ -241,7 +237,7 @@ rpgkit update
 
 ## 今後の機能
 
-- **よりシンプルな生成コマンド:** 現在の多段階の生成フローを、`/rpgkit.generate_repo`、`/rpgkit.generate_feature`、`/rpgkit.plan` などのより少ないコマンドにまとめます。
+- **よりシンプルな生成コマンド:** 現在の多段階の生成フローを、`/rpgkit.generate_repo` や `/rpgkit.generate_feature` などのより少ないコマンドにまとめます。`/rpgkit.plan` は 0.1.4 でリリース済みです。
 - **多言語サポート:** Go、C++、Rust、JavaScript/TypeScript などのサポートを追加します。
 - **より多くのプラットフォーム連携:** さまざまなシステム上の異なる AI コーディングエージェントについて、CLI と VS Code 拡張ワークフローを横断して RPG-Kit をサポートします。
 

RPG-Kit/README.ko-KR.md

@@ -137,11 +137,7 @@ uvx --from "git+https://github.com/microsoft/RPG-ZeroRepo.git#subdirectory=RPG-K
    /rpgkit.feature_build
    /rpgkit.feature_refactor
    [Optional] /rpgkit.feature_edit <edit instructions>
-   /rpgkit.build_skeleton
-   /rpgkit.build_data_flow
-   /rpgkit.design_base_classes
-   /rpgkit.design_interfaces
-   /rpgkit.plan_tasks
+   /rpgkit.plan
    /rpgkit.code_gen
    [Optional] /rpgkit.rpg_edit <edit instructions>
    ```
@@ -241,7 +237,7 @@ rpgkit update
 
 ## 예정된 기능
 
-- **더 간단한 생성 커맨드:** 현재의 다단계 생성 흐름을 `/rpgkit.generate_repo`, `/rpgkit.generate_feature`, `/rpgkit.plan` 등 더 적은 커맨드로 통합합니다.
+- **더 간단한 생성 커맨드:** 현재의 다단계 생성 흐름을 `/rpgkit.generate_repo`, `/rpgkit.generate_feature` 등 더 적은 커맨드로 통합합니다. `/rpgkit.plan` 은 0.1.4 에서 출시되었습니다.
 - **다국어 지원:** Go, C++, Rust, JavaScript/TypeScript 등을 추가로 지원합니다.
 - **더 많은 플랫폼 통합:** 다양한 시스템에서 서로 다른 AI 코딩 에이전트의 CLI 및 VS Code 확장 워크플로에 걸쳐 RPG-Kit을 지원합니다.
 

RPG-Kit/README.md

@@ -137,11 +137,7 @@ Use this path when you want RPG-Kit to turn requirements into a new codebase.
    /rpgkit.feature_build
    /rpgkit.feature_refactor
    [Optional] /rpgkit.feature_edit <edit instructions>
-   /rpgkit.build_skeleton
-   /rpgkit.build_data_flow
-   /rpgkit.design_base_classes
-   /rpgkit.design_interfaces
-   /rpgkit.plan_tasks
+   /rpgkit.plan
    /rpgkit.code_gen
    [Optional] /rpgkit.rpg_edit <edit instructions>
    ```
@@ -241,7 +237,7 @@ rpgkit update
 
 ## Upcoming Features
 
-- **Simpler generation commands:** merge the current multi-step generation flow into fewer commands, such as `/rpgkit.generate_repo`, `/rpgkit.generate_feature`, and `/rpgkit.plan`.
+- **Simpler generation commands:** merge the current multi-step generation flow into fewer commands, such as `/rpgkit.generate_repo` and `/rpgkit.generate_feature`. `/rpgkit.plan` has shipped in 0.1.4.
 - **Multi-language support:** add support for Go, C++, Rust, JavaScript/TypeScript, and more.
 - **More platform integrations:** support RPG-Kit across CLI and VS Code extension workflows for different AI coding agents on different systems.
 

RPG-Kit/README.zh-CN.md

@@ -137,11 +137,7 @@ uvx --from "git+https://github.com/microsoft/RPG-ZeroRepo.git#subdirectory=RPG-K
    /rpgkit.feature_build
    /rpgkit.feature_refactor
    [Optional] /rpgkit.feature_edit <edit instructions>
-   /rpgkit.build_skeleton
-   /rpgkit.build_data_flow
-   /rpgkit.design_base_classes
-   /rpgkit.design_interfaces
-   /rpgkit.plan_tasks
+   /rpgkit.plan
    /rpgkit.code_gen
    [Optional] /rpgkit.rpg_edit <edit instructions>
    ```
@@ -241,7 +237,7 @@ rpgkit update
 
 ## 即将推出的功能
 
-- **更简化的生成命令**：把当前多步骤的生成流程合并为更少的命令，例如 `/rpgkit.generate_repo`、`/rpgkit.generate_feature` 和 `/rpgkit.plan`。
+- **更简化的生成命令**：把当前多步骤的生成流程合并为更少的命令，例如 `/rpgkit.generate_repo` 和 `/rpgkit.generate_feature`。`/rpgkit.plan` 已在 0.1.4 中发布。
 - **多语言支持**：增加对 Go、C++、Rust、JavaScript/TypeScript 等的支持。
 - **更多平台集成**：在不同系统上跨 CLI 和 VS Code 扩展工作流支持不同的 AI 编码智能体。
 

RPG-Kit/docs/commands.md

@@ -1,6 +1,6 @@
 # /rpgkit Commands Reference
 
-RPG-Kit provides 13 slash commands that work in three paths:
+RPG-Kit provides 15 slash commands that work in three paths:
 
 - **Forward pipeline:** Requirements → Repository Planning Graph (RPG) → Code
 - **Reverse encoder:** Existing code → RPG
@@ -14,21 +14,31 @@ RPG-Kit provides 13 slash commands that work in three paths:
 
 | Command | Description |
 | ------- | ----------- |
+| `/rpgkit.feature_construct <desc>` | Run Phase 1 feature specification, build, and refactor in one step — recommended |
 | `/rpgkit.feature_spec <desc>` | Create structured feature specifications from user input or `docs/` files |
 | `/rpgkit.feature_build` | Generate and expand the feature tree from specifications |
 | `/rpgkit.feature_refactor` | Refactor feature tree into modular component architecture |
 | `/rpgkit.feature_edit <instr>` | Edit feature tree nodes before skeleton planning — optional |
 
+> `/rpgkit.feature_construct` is the simplest way to drive Phase 1 end-to-end. Use
+> the individual Phase 1 commands only when you want to debug or re-run a
+> specific stage.
+
 ### Phase 2: RPG Construction and Planning
 
 | Command | Description |
 | ------- | ----------- |
+| `/rpgkit.plan` | Run all five Phase-2 stages in one step with automatic resume — recommended |
 | `/rpgkit.build_skeleton` | Build repository file skeleton from component architecture; creates `.rpgkit/data/rpg.json` |
 | `/rpgkit.build_data_flow` | Build inter-component data flow DAG and update the RPG |
 | `/rpgkit.design_base_classes` | Design shared base classes and data structures |
 | `/rpgkit.design_interfaces` | Design function/class interfaces with type hints and docstrings |
 | `/rpgkit.plan_tasks` | Plan dependency-ordered implementation task batches |
 
+> `/rpgkit.plan` is the simplest way to drive Phase 2 end-to-end. Use
+> the individual commands above only when you want to debug or
+> re-run a specific stage.
+
 ### Phase 3: Code Generation and Surgical Edits
 
 | Command | Description |
@@ -49,6 +59,51 @@ Both directions produce the same RPG structure at `.rpgkit/data/rpg.json`, enabl
 
 ## Phase 1: Feature Specification
 
+### `/rpgkit.feature_construct`
+
+Run the full Phase 1 pipeline (`feature_spec` → `feature_build` → `feature_refactor`) in one step. This is the recommended entry point for creating the feature tree that feeds `/rpgkit.plan`.
+
+**Input modes:**
+
+- **Direct input:** provide requirements after the command.
+- **Auto-detect:** omit input to use existing `docs/*.md` files automatically.
+- **Inline prompt:** if neither direct input nor usable docs exist, the command asks for requirements and then continues in the same flow.
+
+**Output:** every artifact produced by the three individual Phase 1 commands — `.rpgkit/data/feature_spec/`, `feature_spec.json`, `feature_build.json`, and `feature_tree.json`.
+
+**Process:**
+
+1. **Probe progress** — runs `rpgkit script feature_construct.py --check-only --json` to see which Phase 1 stages already have valid artifacts.
+2. **Generate requirements artifacts when needed** — follows the `/rpgkit.feature_spec` workflow for direct text or `docs/*.md` sources.
+3. **Run/resume** — executes `rpgkit script feature_construct.py`, skipping completed stages and cascading downstream rebuilds when an upstream stage reruns.
+4. **Optional expansion** — after completion, the user can expand features through the existing `feature_build --mode suggest-directions` and `--mode step2 --direction <indices>` flow; refactor is rerun afterward so `feature_tree.json` stays aligned.
+
+**CLI flags forwarded after `$ARGUMENTS`:**
+
+- `--check-only` — show Phase 1 progress without modifying artifacts or running stages.
+- `--json` — with `--check-only`, emit the progress report as JSON.
+- `--force` — rebuild all Phase 1 stages.
+- `--dry-run` — print the commands that would run without modifying artifacts.
+- `--verbose` — forward native verbose logging flags.
+- `--no-trajectory` — disable trajectory recording where supported.
+- `--max-iter-refactor N` — forward to `feature_refactor.py` as `--max-iterations N`.
+- `--review-threshold N` — forward to `feature_build.py --mode step1`.
+- `--review-max-iterations N` — forward to `feature_build.py --mode step1`.
+
+Use `--` to separate options from requirement text:
+
+```text
+/rpgkit.feature_construct --check-only
+/rpgkit.feature_construct --check-only --json
+/rpgkit.feature_construct --review-threshold 99 -- Build a CLI tool for managing Docker containers
+/rpgkit.feature_construct Build a CLI tool for managing Docker containers
+/rpgkit.feature_construct                  # Auto-detect docs/ files
+```
+
+**Next step:** `/rpgkit.plan` is the default handoff after Phase 1. If the final tree needs small adjustments, run `/rpgkit.feature_edit <instructions>`. The granular Phase 1 commands remain available for debug and surgical reruns; `/rpgkit.build_skeleton` is a Phase 2 granular fallback, not the default next step.
+
+---
+
 ### `/rpgkit.feature_spec`
 
 Create structured feature specifications from user input or documentation files.
@@ -159,6 +214,69 @@ Edit feature tree nodes before repository planning begins.
 
 ## Phase 2: RPG Construction and Planning
 
+### `/rpgkit.plan`
+
+Run the full Phase-2 pipeline (`build_skeleton` → `build_data_flow` →
+`design_base_classes` → `design_interfaces` → `plan_tasks`) in one
+step. This is the recommended entry point for Phase 2.
+
+**Input:** `~/.rpgkit/workspaces/<workspace-id>/data/feature_tree.json` (produced by
+`/rpgkit.feature_construct` or `/rpgkit.feature_refactor`)
+
+**Output:** every artifact produced by the five individual commands —
+`skeleton.json`, `data_flow.json`, `base_classes.json`,
+`interfaces.json`, `tasks.json`, plus `rpg.json` and the
+`data_flow_viz.html` visualization.
+
+**Process:**
+
+1. **Probe progress** — runs `rpgkit script plan.py --check-only --json`
+   to see which stages already have valid artifacts.
+2. **Decide** — based on the probe result, the command prompts you
+   **once** with one of three options:
+    - All five stages already done → `Overwrite` or `Exit`.
+    - Partial progress (some stages done) → `Continue`, `Restart`, or `Exit`.
+    - Fresh workspace → no prompt; runs the full pipeline.
+3. **Run** — executes the chosen mode through `rpgkit script plan.py`.
+   Each stage's stdout is streamed live and also written to a per-stage
+   log under `~/.rpgkit/workspaces/<workspace-id>/logs/`.
+4. **Verify** — after every stage's build script, the corresponding
+   `check_*.py` script re-runs to validate the produced artifact. If
+   verification fails the pipeline stops and prints recovery hints.
+
+**Resume semantics:** the command treats `type == "update"` from each
+`check_*.py` as the source of truth for "this stage is done". If you
+press Ctrl-C halfway through, running `/rpgkit.plan` again automatically
+resumes from the first not-done stage. When any earlier stage is
+re-run, every downstream stage is rebuilt too so artifacts never drift
+apart.
+
+**CLI flags forwarded after `$ARGUMENTS`:**
+
+- `--force` — discard existing artifacts and rebuild every stage.
+- `--max-iter-skeleton N`, `--max-iter-data-flow N`,
+  `--max-iter-base-classes N`, `--max-iter-interfaces N` —
+  override iteration counts for the corresponding stage.
+- `--verbose` — forward `--verbose` to every sub-script.
+- `--no-trajectory` — forward `--no-trajectory` where supported.
+
+**Examples:**
+
+```text
+/rpgkit.plan
+/rpgkit.plan --verbose
+/rpgkit.plan --force                    # rebuild everything
+/rpgkit.plan --max-iter-skeleton 15
+```
+
+To inspect progress without running anything:
+
+```bash
+rpgkit script plan.py --check-only
+```
+
+---
+
 ### `/rpgkit.build_skeleton`
 
 Build the repository file skeleton from the component architecture. This is where the forward pipeline first creates the RPG.
@@ -462,10 +580,10 @@ All intermediate data is stored in `.rpgkit/data/`:
 
 | File | Produced by | Description |
 | ---- | ----------- | ----------- |
-| `feature_spec/` | `feature_spec` | Evidence and feature specification documents |
-| `feature_spec.json` | `feature_spec` | Structured feature specification |
-| `feature_build.json` | `feature_build` | Expanded feature tree |
-| `feature_tree.json` | `feature_refactor` / `feature_edit` | Component architecture |
+| `feature_spec/` | `feature_construct` / `feature_spec` | Evidence and feature specification documents |
+| `feature_spec.json` | `feature_construct` / `feature_spec` | Structured feature specification |
+| `feature_build.json` | `feature_construct` / `feature_build` | Expanded feature tree |
+| `feature_tree.json` | `feature_construct` / `feature_refactor` / `feature_edit` | Component architecture |
 | `skeleton.json` | `build_skeleton` | File skeleton |
 | `skeleton_summary.txt` | `build_skeleton` | Human-readable skeleton summary |
 | `rpg.json` | `build_skeleton` / `encode`, then updated by later commands | Repository Planning Graph |

RPG-Kit/scripts/check_base_classes.py

@@ -108,14 +108,14 @@ def inspect_state(base_classes_path: Path) -> Dict[str, Any]:
     """Inspect current state and determine action needed.
 
     Returns dict with:
-    - state: "error" | "init" | "update"
+    - type: "error" | "init" | "update"
     - message: description
     - details: additional info
     """
     # Check if base_classes.json exists
     if not base_classes_path.exists():
         return {
-            "state": "init",
+            "type": "init",
             "message": "base_classes.json not found - need to run design_base_classes",
             "details": {}
         }
@@ -126,15 +126,15 @@ def inspect_state(base_classes_path: Path) -> Dict[str, Any]:
             data = json.load(f)
     except json.JSONDecodeError as e:
         return {
-            "state": "error",
+            "type": "error",
             "message": f"Invalid JSON in base_classes.json: {e}",
             "details": {}
         }
 
     # Check for error field
     if "error" in data:
         return {
-            "state": "error",
+            "type": "error",
             "message": f"Base classes has error: {data['error']}",
             "details": {}
         }
@@ -143,7 +143,7 @@ def inspect_state(base_classes_path: Path) -> Dict[str, Any]:
     is_valid, errors = validate_base_classes_structure(data)
     if not is_valid:
         return {
-            "state": "error",
+            "type": "error",
             "message": "Base classes structure or syntax is invalid",
             "details": {"errors": errors}
         }
@@ -161,7 +161,7 @@ def inspect_state(base_classes_path: Path) -> Dict[str, Any]:
     ds_file_paths = [ds.get("file_path", "") for ds in data_structures if ds.get("file_path")]
 
     return {
-        "state": "update",
+        "type": "update",
         "message": "Base classes are valid",
         "details": {
             "file_count": len(base_classes),
@@ -178,7 +178,7 @@ def inspect_state(base_classes_path: Path) -> Dict[str, Any]:
 
 def print_state(result: Dict[str, Any]) -> None:
     """Print state information."""
-    state = result["state"]
+    state = result["type"]
     message = result["message"]
     details = result.get("details", {})
 
@@ -259,7 +259,7 @@ def main():
     result = inspect_state(args.input)
 
     # In verbose mode, include raw base_classes data
-    if args.verbose and result.get("state") == "update":
+    if args.verbose and result.get("type") == "update":
         base_classes_data = load_json(args.input)
         if base_classes_data:
             result["base_classes"] = base_classes_data.get("base_classes", [])
@@ -273,7 +273,7 @@ def main():
         print_state(result)
 
     # Return exit code based on state
-    if result["state"] == "error":
+    if result["type"] == "error":
         return 1
     return 0