chore(unity): 固定Unity CLI Loop本地包并移除UnitySkills依赖

Client/Packages/io.github.hatayama.uloopmcp/Editor/Api/McpTools/ReplayInput/Skill/SKILL.md

@@ -0,0 +1,60 @@
+---
+name: uloop-replay-input
+description: "Replay recorded PlayMode keyboard and mouse input. Use for exact gameplay reproduction, E2E runs, or consistent demos from JSON recordings."
+---
+
+# uloop replay-input
+
+Replay recorded keyboard and mouse input during PlayMode. Loads a JSON recording and injects input frame-by-frame via Input System with zero CLI overhead. Supports looping and progress monitoring.
+
+## Usage
+
+```bash
+# Start replay (auto-detect latest recording)
+uloop replay-input --action Start
+
+# Start replay with specific file
+uloop replay-input --action Start --input-path scripts/my-play.json
+
+# Start replay with looping
+uloop replay-input --action Start --loop true
+
+# Check replay progress
+uloop replay-input --action Status
+
+# Stop replay
+uloop replay-input --action Stop
+```
+
+## Parameters
+
+| Parameter | Type | Default | Description |
+|-----------|------|---------|-------------|
+| `--action` | enum | `Start` | `Start`, `Stop`, `Status` |
+| `--input-path` | string | auto | JSON path. Auto-detects latest in `.uloop/outputs/InputRecordings/` |
+| `--show-overlay` | boolean | `true` | Show replay progress overlay |
+| `--loop` | boolean | `false` | Loop continuously |
+
+## Deterministic Replay
+
+Replay injects the exact same input frame-by-frame, but the game must also be deterministic to produce identical results. If replay output must be compared across runs, read [references/deterministic-replay.md](references/deterministic-replay.md) before interpreting failures.
+
+## Prerequisites
+
+- Unity must be in **PlayMode**
+- **Input System package** must be installed (`com.unity.inputsystem`)
+- Use this only when the project already uses the New Input System.
+
+## Output
+
+Returns JSON with:
+- `Success`: Whether the operation succeeded
+- `Message`: Status message
+- `Action`: Echoes which action was executed (`Start`, `Stop`, or `Status`)
+- `InputPath`: Path to recording file (nullable string; populated on `Start` only)
+- `CurrentFrame`: Current replay frame index (nullable int)
+- `TotalFrames`: Total frames in the recording (nullable int)
+- `Progress`: Replay progress (nullable float in 0.0 – 1.0)
+- `IsReplaying`: Whether replay is currently active (nullable bool)
+
+These are the only eight fields. There is no `LoopCount`, `ElapsedSeconds`, `OverlayVisible`, or per-frame inspection data in the response.

Client/Packages/io.github.hatayama.uloopmcp/Editor/Api/McpTools/ReplayInput/Skill/SKILL.md.meta

@@ -0,0 +1,7 @@
+fileFormatVersion: 2
+guid: d413553edea5f45158e4adf599a059df
+TextScriptImporter:
+  externalObjects: {}
+  userData: 
+  assetBundleName: 
+  assetBundleVariant: 

Client/Packages/io.github.hatayama.uloopmcp/Editor/Api/McpTools/ReplayInput/Skill/references.meta

@@ -0,0 +1,8 @@
+fileFormatVersion: 2
+guid: b8ef7887274da4fa5bfcad96ea3d6281
+folderAsset: yes
+DefaultImporter:
+  externalObjects: {}
+  userData: 
+  assetBundleName: 
+  assetBundleVariant: 

Client/Packages/io.github.hatayama.uloopmcp/Editor/Api/McpTools/ReplayInput/Skill/references/deterministic-replay.md

@@ -0,0 +1,15 @@
+# Deterministic Replay
+
+Replay injects input frame-by-frame, so the game must produce identical results given identical input on each run. The following patterns break determinism and should be avoided in replay-targeted code:
+
+| Avoid | Use Instead | Why |
+|-------|-------------|-----|
+| `Time.deltaTime` for movement | Fixed per-frame constant (e.g. `MOVE_SPEED = 0.1f`) | Delta time varies between runs even at the same target frame rate |
+| `Random.Range()` / `UnityEngine.Random` | Seeded random (`new System.Random(fixedSeed)`) or remove randomness | Random sequences differ between runs |
+| `Rigidbody` / Physics simulation | Kinematic movement via `Transform.Translate` | Physics is non-deterministic across runs |
+| `WaitForSeconds(n)` in coroutines | `WaitForEndOfFrame` or frame counting | Real-time waits depend on frame timing |
+| `Time.time` / `Time.realtimeSinceStartup` | Frame counter (`Time.frameCount - startFrame`) | Time values drift between runs |
+| `FindObjectsOfType` without sort | `FindObjectsByType(FindObjectsSortMode.InstanceID)` | Iteration order is non-deterministic |
+| `async/await` with `Task.Delay` | Frame-based waiting | Real-time delays are non-deterministic |
+
+Set `Application.targetFrameRate = 60` or another fixed target to reduce frame timing variance. See `InputReplayVerificationController` for a complete example of deterministic game logic.

Client/Packages/io.github.hatayama.uloopmcp/Editor/Api/McpTools/ReplayInput/Skill/references/deterministic-replay.md.meta

@@ -0,0 +1,7 @@
+fileFormatVersion: 2
+guid: af36ab88f77db43a4b6b3ef19dd4432b
+TextScriptImporter:
+  externalObjects: {}
+  userData: 
+  assetBundleName: 
+  assetBundleVariant: