docs: add per-tool API documentation and fix parity gaps

- Create 53 per-tool VitePress documentation pages with accurate
  parameters from Zod schemas, example requests, and response formats
- Add root llms.txt for LLM-friendly repo browsing
- Fix OpenAPI spec: add auth and 422 error schemas to
  edit-metadata/inspect and strip-metadata/inspect sub-routes
- Fix tool count inconsistency (52 -> 53) across landing site,
  e2e tests, and local docs
- Rename color-adjustments.ts to adjust-colors.ts to match tool ID
- Update VitePress sidebar with all 8 tool categories and top nav

apps/docs/tools/restore-photo.md

@@ -0,0 +1,89 @@
+# Photo Restoration
+
+Fix scratches, tears, and damage on old photos using a multi-step AI pipeline. Combines scratch repair, face enhancement, denoising, and optional colorization.
+
+## API Endpoint
+
+`POST /api/v1/tools/restore-photo`
+
+**Processing:** Asynchronous (returns 202, poll `/api/v1/jobs/{jobId}/progress` for status via SSE)
+
+**Model bundle:** `photo-restoration` (800 MB - 1 GB)
+
+## Parameters
+
+| Parameter | Type | Required | Default | Description |
+|-----------|------|----------|---------|-------------|
+| file | file | Yes | - | Image file (multipart) |
+| scratchRemoval | boolean | No | `true` | Remove scratches and surface damage |
+| faceEnhancement | boolean | No | `true` | Enhance faces in the restored photo |
+| fidelity | number | No | `0.7` | Face enhancement fidelity (0-1). Higher values preserve original features more |
+| denoise | boolean | No | `true` | Apply denoising to the restored result |
+| denoiseStrength | number | No | `25` | Denoising strength (0-100) |
+| colorize | boolean | No | `false` | Colorize the restored photo (for grayscale images) |
+| colorizeStrength | number | No | `85` | Colorization intensity (0-100) |
+
+## Example Request
+
+```bash
+curl -X POST http://localhost:13490/api/v1/tools/restore-photo \
+  -F "file=@damaged-old-photo.jpg" \
+  -F 'settings={"scratchRemoval":true,"faceEnhancement":true,"fidelity":0.6,"colorize":true}'
+```
+
+## Response
+
+### Initial Response (202 Accepted)
+
+```json
+{
+  "jobId": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
+  "async": true
+}
+```
+
+### Progress (SSE at `/api/v1/jobs/{jobId}/progress`)
+
+```
+event: progress
+data: {"phase":"processing","stage":"Removing scratches...","percent":30}
+```
+
+```
+event: progress
+data: {"phase":"processing","stage":"Enhancing faces...","percent":60}
+```
+
+### Final Result (via SSE)
+
+```json
+{
+  "phase": "complete",
+  "percent": 100,
+  "result": {
+    "jobId": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
+    "downloadUrl": "/api/v1/download/{jobId}/damaged-old-photo_restored.jpg",
+    "previewUrl": "/api/v1/download/{jobId}/preview.webp",
+    "originalSize": 200000,
+    "processedSize": 350000,
+    "width": 1200,
+    "height": 900,
+    "steps": ["scratch_removal", "face_enhancement", "denoise", "colorize"],
+    "scratchCoverage": 12.5,
+    "facesEnhanced": 2,
+    "isGrayscale": true,
+    "colorized": true
+  }
+}
+```
+
+## Notes
+
+- Requires the `photo-restoration` model bundle to be installed (800 MB - 1 GB).
+- The pipeline runs multiple AI steps sequentially: scratch repair, face enhancement (GFPGAN), denoising, and optionally colorization.
+- The `steps` array in the result shows which processing steps were actually executed.
+- `scratchCoverage` is an estimated percentage of the image area that had scratch damage.
+- `fidelity` controls how strongly faces are enhanced vs. preserving the original appearance. Lower values produce more aggressive enhancement; higher values are more conservative.
+- The `colorize` option automatically detects if the image is grayscale. The `isGrayscale` flag in the result confirms this detection.
+- Output format matches the input format automatically.
+- Supports HEIC/HEIF, RAW, TGA, PSD, EXR, HDR, and AVIF input formats via automatic decoding.