From 2dc86e53212ab322c41a212a1e16564c2619136b Mon Sep 17 00:00:00 2001
From: Rene Zander <renezander030@gmail.com>
Date: Fri, 25 Sep 2026 06:24:46 +0000
Subject: [PATCH] release: prepare v0.26.0 workflow upgrades

---
 CHANGELOG.md                        |  12 +
 README.md                           |   2 +
 README.zh-CN.md                     |   2 +
 docs/command-reference.json         | 387 +++++++++++++++++++++++++++-
 docs/command-reference.md           |   7 +-
 docs/command-reference.zh-CN.md     |  13 +-
 package-lock.json                   |   4 +-
 package.json                        |   2 +-
 python/capcut/__init__.py           |   2 +-
 python/pyproject.toml               |   2 +-
 src/align.ts                        |  75 +++++-
 src/caption.ts                      | 107 +++++++-
 src/command-specs.ts                |  77 +++++-
 src/factory.ts                      |  41 +++
 src/index.ts                        | 184 ++++++++++++-
 src/interchange.ts                  | 179 +++++++++++--
 src/lint.ts                         |  62 +++++
 src/render.ts                       |  63 ++++-
 test/caption-script.test.mjs        |  52 +++-
 test/caption-stream-reveal.test.mjs | 130 ++++++++++
 test/caption.test.mjs               |   7 +
 test/edit.test.mjs                  |  44 ++++
 test/import-timeline.test.mjs       |  59 ++++-
 test/lint-frame-grid.test.mjs       |  77 ++++++
 test/make-preset.test.mjs           |  42 +++
 test/remove.test.mjs                |  28 ++
 test/render.test.mjs                |  45 ++++
 27 files changed, 1614 insertions(+), 91 deletions(-)
 create mode 100644 test/caption-stream-reveal.test.mjs
 create mode 100644 test/lint-frame-grid.test.mjs

diff --git a/CHANGELOG.md b/CHANGELOG.md
index b959d16..057ff57 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -4,8 +4,20 @@ All notable changes to capcut-cli are documented here. The format follows [Keep
 
 ## [Unreleased]
 
+## [0.26.0] — 2026-09-25
+
 ### Added
 
+- `lint --frame-grid` validates every segment start and end against the draft's fps. `lint --frame-grid --fix` snaps the two boundaries and derives the duration from them, avoiding the one-microsecond overlaps that independent start/duration rounding can create; an unchanged 1x source range follows the repaired duration.
+- `caption --script` now aligns Chinese and Japanese at character granularity, including when Whisper returns several characters as one timed token. `--min-script-match <0..1>` turns the alignment report into a write gate: a mismatched script is refused before the draft changes.
+- `caption --audio-stream <n>` selects a zero-based audio stream from a multi-stream input. FFmpeg extracts that stream to a temporary mono 16 kHz WAV before Whisper runs; `--ffmpeg-cmd` selects the binary and the result reports `audio_stream` while keeping `source_audio` pointed at the original container.
+- Boundary-safe ripple editing: `remove --ripple` closes the removed span across every track and refuses before writing when another segment crosses it; `shift-all --from <time>` moves only segments at or after an exact boundary and likewise refuses a boundary that cuts through a segment.
+- Large proxy renders automatically pass filter chains longer than 8 KiB through FFmpeg's filter-script input instead of the process command line, then remove the temporary script after the render. Dry-run plans expose the script path and content without writing it.
+- `import-timeline` flattens nested OTIO `Timeline.1`, `Stack.1`, and `Track.1` sequences into editable CapCut tracks while preserving parent gaps, offsets, durations, and nested caption markers. Unsupported effects and items remain explicitly reported.
+- `render --crf <0..51>` controls constant quality (default 28); `render --video-bitrate <rate>` selects a target bitrate such as `2500k` or `4M`. The two modes are mutually exclusive and the render plan reports which one it uses.
+- `caption --word-reveal` writes progressive, word-timed caption prefixes (`one` → `one two` → `one two three`) without flattening them into pixels. It is mutually exclusive with karaoke.
+- `restyle <project> --preset <file> [--track-name <name>]` applies one text-style preset atomically to every text segment or to one caption track. Explicit style flags override the preset and the result reports affected track/segment counts.
+
 - `doctor` reports what each draft store holds — for every default CapCut/JianYing project directory it finds (or the one folder named with the new `--drafts <dir>`), a `draft-store` check counts the projects as readable, markerless, encrypted or unreadable. A JianYing 6.0+ store, where every project the app wrote is an encrypted payload, is now named once and up front (warn) with what still works — `init`, `quickstart` and `compile` build plaintext drafts from the bundled template — instead of being discovered one failed command at a time. Same classification as the `template.store` report of `init`/`quickstart`/`compile`. `capcut doctor --drafts <dir>` also makes the check usable on a machine without the app, and in CI.
 - `examples/short-video-narration.md` (+ zh-CN) — silent clip → 9:16 draft with a TTS voiceover and script-accurate captions, as four commands (`quickstart --ratio 9:16` → `tts --text-file` → `caption --from-segment --script` → `lint`) and as one script, `examples/scripts/narrate-short.sh`. `examples/scripts/edge-tts-wav.sh` bridges edge-tts (MP3 only) to the WAV `tts` expects at `{out}`; any other engine plugs in through `--tts-cmd`. The vision-model step that writes the script is optional and stays outside the CLI: the script is a text file.
 - `python/` — a thin Python client, published to PyPI as `capcut` (`pip install capcut`). `capcut.run(cmd, *args, **flags)` spawns the CLI once without a shell and returns the JSON it prints; keyword arguments become flags (`font_size=16` → `--font-size 16`), positionals pass through as single argv tokens, a non-zero exit raises `CommandError` with `status` and the CLI's JSON. `capcut.serve(jobs)` feeds the stateless JSONL queue and returns one result per job. `capcut.describe()`, `capcut.doctor()`, `capcut.version()`. Pure Python, no dependencies, Python ≥ 3.9; the binary is found on PATH or through `CAPCUT_CLI`. Not part of the npm tarball.
diff --git a/README.md b/README.md
index 6944621..86f90b5 100644
--- a/README.md
+++ b/README.md
@@ -105,6 +105,8 @@ The host reads a draft and passes its JSON as tool input. The component itself h
 
 ## Release notes
 
+> **New in v0.26.0:** exact frame-grid lint/fix; character-level Chinese/Japanese script alignment with an optional match gate; explicit caption audio-stream selection; safe ripple delete and boundary shifts; scalable FFmpeg filter scripts; nested OTIO import; CRF/bitrate proxy controls; progressive word-reveal captions; and atomic whole-track `restyle`. Full details in the [changelog](./CHANGELOG.md).
+
 > **New in v0.25.0:** `caption` follows the transcript's script. Whisper's "words" for Chinese and Japanese are single characters or short tokens, so the Latin defaults (four words per cue, joined with spaces) produced fragments with spaces between the characters; cues are now joined without spaces and bounded by characters alone, at the width `lint` holds captions to (zh 16, ja 13, ko 16), and the result reports `caption_script`. An explicit `--max-words` / `--max-chars` still wins. Full details in the [changelog](./CHANGELOG.md).
 
 > **New in v0.24.0:** captions in Chinese, Japanese and Korean are held to their own limits — `lint` flags a 32-character Chinese line and a 15 chars/s cue that the Latin defaults (42, 20) let through, and `--fix` re-wraps between characters (zh 16/9, ja 13/4, ko 16/12; an explicit `--max-chars` / `--max-cps` still applies everywhere). On a JianYing 6.0+ drafts folder, where every app-written project is encrypted, `init` / `quickstart` / `compile` now say that none could seed the new draft (`template.store`, a WARNING) and `lint` reports `template-unverified-store` instead of nothing. Plus a one-command agent install: `npx skills add renezander030/capcut-cli`. Full details in the [changelog](./CHANGELOG.md).
diff --git a/README.zh-CN.md b/README.zh-CN.md
index e7840a4..291014f 100644
--- a/README.zh-CN.md
+++ b/README.zh-CN.md
@@ -82,6 +82,8 @@ Claude Code 也可以把它作为插件加载：
 
 ## 发布说明
 
+> **v0.26.0 新增：** 精确帧网格检查/修复；中日文按字对齐脚本并可设置匹配率门槛；字幕音轨选择；安全波纹删除和边界平移；大型 FFmpeg 滤镜脚本；嵌套 OTIO 导入；CRF/码率预览控制；逐词显现字幕；以及整轨原子化 `restyle`。完整说明见[更新日志](./CHANGELOG.md)。
+
 > **v0.25.0 新增：** `caption` 按转写文本的文字来分句。Whisper 对中文、日文给出的"词"是单个字或很短的片段，按拉丁默认（每句 4 词、用空格连接）会生成字与字之间带空格的碎片；现在中日文按字直接连接、只按字数上限分句，上限就是 `lint` 对字幕的行宽（zh 16、ja 13、ko 16），结果里会报告 `caption_script`。显式传入的 `--max-words` / `--max-chars` 仍然优先。完整说明见[更新日志](./CHANGELOG.md)。
 
 > **v0.24.0 新增：** 中文、日文、韩文字幕按各自的规范检查 —— `lint` 会指出 32 字的中文单行和每秒 15 字的字幕（拉丁默认的 42 字 / 每秒 20 字会放过它们），`--fix` 按字重新折行（zh 16/9、ja 13/4、ko 16/12；显式传入 `--max-chars` / `--max-cps` 仍对所有文字生效）。在剪映 6.0+ 的草稿目录里（应用写出的项目全部加密），`init` / `quickstart` / `compile` 现在会明确说明没有任何项目可作为种子（`template.store` 与 WARNING），`lint` 会报告 `template-unverified-store` 而不是沉默。另外，一条命令即可把它装进 Agent：`npx skills add renezander030/capcut-cli`。完整说明见[更新日志](./CHANGELOG.md)。
diff --git a/docs/command-reference.json b/docs/command-reference.json
index a93e1e6..9e056ef 100644
--- a/docs/command-reference.json
+++ b/docs/command-reference.json
@@ -1,6 +1,6 @@
 {
   "name": "capcut-cli",
-  "version": "0.25.0",
+  "version": "0.26.0",
   "schema_version": 2,
   "description": "Edit CapCut/JianYing draft_content.json directly. JSON in, JSON out.",
   "global_flags": [
@@ -195,6 +195,15 @@
           "required": false,
           "description": "Mechanically repair fixable issues and write the draft."
         },
+        {
+          "name": "frame_grid",
+          "flags": [
+            "--frame-grid"
+          ],
+          "type": "boolean",
+          "required": false,
+          "description": "Check target ranges against draft.fps; --fix snaps boundaries and derives duration."
+        },
         {
           "name": "no_probe",
           "flags": [
@@ -388,8 +397,8 @@
     },
     {
       "name": "shift-all",
-      "summary": "Shift all segments (optionally on one --track) by an offset.",
-      "usage": "capcut shift-all <project> <offset> [--track <type>]",
+      "summary": "Shift segments by an offset, optionally filtering by track and start time.",
+      "usage": "capcut shift-all <project> <offset> [--track <type>] [--from <time>]",
       "positionals": [
         {
           "name": "project",
@@ -405,6 +414,11 @@
           "name": "type",
           "type": "string",
           "required": false
+        },
+        {
+          "name": "time",
+          "type": "time",
+          "required": false
         }
       ],
       "options": [
@@ -417,6 +431,15 @@
           "type": "string",
           "required": false,
           "description": "Track or material type filter."
+        },
+        {
+          "name": "from_time",
+          "flags": [
+            "--from"
+          ],
+          "type": "time",
+          "required": false,
+          "description": "Shift only segments starting at or after this boundary; refuses a segment crossing it."
         }
       ],
       "mutates": true,
@@ -720,7 +743,7 @@
     },
     {
       "name": "import-timeline",
-      "summary": "Import OpenTimelineIO JSON (the export-timeline schema set) as a new draft (--out) or append it onto an existing one (--into); unsupported OTIO features are reported, never silent.",
+      "summary": "Import OpenTimelineIO JSON, flattening nested Timeline/Stack/Track sequences, as a new draft (--out) or append it onto an existing one (--into); unsupported OTIO features are reported, never silent.",
       "usage": "capcut import-timeline <file.otio> (--out <new-project> | --into <project>)",
       "positionals": [
         {
@@ -1433,7 +1456,7 @@
     {
       "name": "remove",
       "summary": "Remove a segment, its emptied track, and the materials that orphans.",
-      "usage": "capcut remove <project> <segment-id> [--keep-track] [--keep-materials]",
+      "usage": "capcut remove <project> <segment-id> [--keep-track] [--keep-materials] [--ripple]",
       "positionals": [
         {
           "name": "project",
@@ -1464,6 +1487,15 @@
           "type": "boolean",
           "required": false,
           "description": "Skip the orphan-material sweep (run prune later)."
+        },
+        {
+          "name": "ripple",
+          "flags": [
+            "--ripple"
+          ],
+          "type": "boolean",
+          "required": false,
+          "description": "Close the removed time span across every track; refuses crossing segments."
         }
       ],
       "mutates": true,
@@ -2050,6 +2082,293 @@
         "1": "invalid input, warning, or operation failure"
       }
     },
+    {
+      "name": "restyle",
+      "summary": "Apply one text-style preset atomically to a whole caption track or every text segment.",
+      "usage": "capcut restyle <project> --preset <preset.json> [--track-name <name>] [options]",
+      "positionals": [
+        {
+          "name": "project",
+          "type": "path",
+          "required": true
+        },
+        {
+          "name": "name",
+          "type": "string",
+          "required": false
+        }
+      ],
+      "options": [
+        {
+          "name": "track_name",
+          "flags": [
+            "--track-name"
+          ],
+          "type": "string",
+          "required": false,
+          "description": "Target track name."
+        },
+        {
+          "name": "font_size",
+          "flags": [
+            "--font-size"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Font size."
+        },
+        {
+          "name": "color",
+          "flags": [
+            "--color"
+          ],
+          "type": "string",
+          "required": false,
+          "description": "Text colour as #RRGGBB."
+        },
+        {
+          "name": "align",
+          "flags": [
+            "--align"
+          ],
+          "type": "enum",
+          "required": false,
+          "description": "Text alignment.",
+          "values": [
+            "0",
+            "1",
+            "2"
+          ]
+        },
+        {
+          "name": "x",
+          "flags": [
+            "--x"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Horizontal position."
+        },
+        {
+          "name": "y",
+          "flags": [
+            "--y"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Vertical position."
+        },
+        {
+          "name": "alpha",
+          "flags": [
+            "--alpha"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Text alpha."
+        },
+        {
+          "name": "vertical",
+          "flags": [
+            "--vertical"
+          ],
+          "type": "boolean",
+          "required": false,
+          "description": "Use vertical text."
+        },
+        {
+          "name": "fixed_width",
+          "flags": [
+            "--fixed-width"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Fixed text-box width."
+        },
+        {
+          "name": "fixed_height",
+          "flags": [
+            "--fixed-height"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Fixed text-box height."
+        },
+        {
+          "name": "shadow",
+          "flags": [
+            "--shadow",
+            "--no-shadow"
+          ],
+          "type": "boolean",
+          "required": false,
+          "description": "Enable or disable shadow."
+        },
+        {
+          "name": "shadow_alpha",
+          "flags": [
+            "--shadow-alpha"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Shadow alpha."
+        },
+        {
+          "name": "shadow_angle",
+          "flags": [
+            "--shadow-angle"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Shadow angle."
+        },
+        {
+          "name": "shadow_color",
+          "flags": [
+            "--shadow-color"
+          ],
+          "type": "string",
+          "required": false,
+          "description": "Shadow colour."
+        },
+        {
+          "name": "shadow_distance",
+          "flags": [
+            "--shadow-distance"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Shadow distance."
+        },
+        {
+          "name": "shadow_smoothing",
+          "flags": [
+            "--shadow-smoothing"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Shadow smoothing."
+        },
+        {
+          "name": "border_width",
+          "flags": [
+            "--border-width"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Border width."
+        },
+        {
+          "name": "border_color",
+          "flags": [
+            "--border-color"
+          ],
+          "type": "string",
+          "required": false,
+          "description": "Border colour."
+        },
+        {
+          "name": "border_alpha",
+          "flags": [
+            "--border-alpha"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Border alpha."
+        },
+        {
+          "name": "bg_color",
+          "flags": [
+            "--bg-color"
+          ],
+          "type": "string",
+          "required": false,
+          "description": "Background colour."
+        },
+        {
+          "name": "bg_alpha",
+          "flags": [
+            "--bg-alpha"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Background alpha."
+        },
+        {
+          "name": "bg_style",
+          "flags": [
+            "--bg-style"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Background style identifier."
+        },
+        {
+          "name": "bg_round_radius",
+          "flags": [
+            "--bg-round-radius"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Background corner radius."
+        },
+        {
+          "name": "bg_width",
+          "flags": [
+            "--bg-width"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Background width."
+        },
+        {
+          "name": "bg_height",
+          "flags": [
+            "--bg-height"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Background height."
+        },
+        {
+          "name": "bg_h_offset",
+          "flags": [
+            "--bg-h-offset"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Background horizontal offset."
+        },
+        {
+          "name": "bg_v_offset",
+          "flags": [
+            "--bg-v-offset"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Background vertical offset."
+        },
+        {
+          "name": "preset",
+          "flags": [
+            "--preset"
+          ],
+          "type": "path",
+          "required": false,
+          "description": "Apply a make-preset style preset; explicit flags override preset values."
+        }
+      ],
+      "mutates": true,
+      "prerequisites": [],
+      "output": {
+        "type": "object",
+        "description": "JSON by default; use -H where supported for human output."
+      },
+      "exit_codes": {
+        "0": "success",
+        "1": "invalid input, warning, or operation failure"
+      }
+    },
     {
       "name": "text-anim",
       "summary": "Add intro/outro/combo text animation.",
@@ -3593,6 +3912,24 @@
           "required": false,
           "description": "Audio input."
         },
+        {
+          "name": "audio_stream",
+          "flags": [
+            "--audio-stream"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Zero-based audio stream to extract from the input container before transcription."
+        },
+        {
+          "name": "ffmpeg_cmd",
+          "flags": [
+            "--ffmpeg-cmd"
+          ],
+          "type": "path",
+          "required": false,
+          "description": "FFmpeg binary used by --audio-stream."
+        },
         {
           "name": "from_segment",
           "flags": [
@@ -3652,6 +3989,24 @@
           "required": false,
           "description": "Create word-highlight caption ranges."
         },
+        {
+          "name": "word_reveal",
+          "flags": [
+            "--word-reveal"
+          ],
+          "type": "boolean",
+          "required": false,
+          "description": "Build each cue progressively, adding one timed word at a time."
+        },
+        {
+          "name": "min_script_match",
+          "flags": [
+            "--min-script-match"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Refuse before writing when caption --script exact-token match ratio is below 0..1."
+        },
         {
           "name": "max_words",
           "flags": [
@@ -3747,7 +4102,8 @@
       ],
       "mutates": true,
       "prerequisites": [
-        "whisper CLI"
+        "whisper CLI",
+        "ffmpeg (only with --audio-stream)"
       ],
       "output": {
         "type": "object",
@@ -5416,6 +5772,25 @@
           "required": false,
           "description": "Video encoder for the proxy (-c:v; default libx264). Hardware encoders like h264_videotoolbox/h264_nvenc/h264_qsv work when the build carries them; validated against `ffmpeg -encoders` before rendering."
         },
+        {
+          "name": "crf",
+          "flags": [
+            "--crf"
+          ],
+          "type": "number",
+          "required": false,
+          "description": "Constant-quality value 0..51 (default 28; mutually exclusive with --video-bitrate).",
+          "default": 28
+        },
+        {
+          "name": "video_bitrate",
+          "flags": [
+            "--video-bitrate"
+          ],
+          "type": "string",
+          "required": false,
+          "description": "Target video bitrate such as 2500k or 4M; switches the proxy from CRF to bitrate mode."
+        },
         {
           "name": "burn_captions",
           "flags": [
diff --git a/docs/command-reference.md b/docs/command-reference.md
index 351942d..d5c06f5 100644
--- a/docs/command-reference.md
+++ b/docs/command-reference.md
@@ -12,7 +12,7 @@
 | `texts` | `capcut texts <project>` | no | List all text/subtitle content. |
 | `set-text` | `capcut set-text <project> <id> <text>` | yes | Change a text segment's content. |
 | `shift` | `capcut shift <project> <id> <offset>` | yes | Shift one segment's timing by an offset (e.g. +0.5s). |
-| `shift-all` | `capcut shift-all <project> <offset> [--track <type>]` | yes | Shift all segments (optionally on one --track) by an offset. |
+| `shift-all` | `capcut shift-all <project> <offset> [--track <type>] [--from <time>]` | yes | Shift segments by an offset, optionally filtering by track and start time. |
 | `speed` | `capcut speed <project> <id> <multiplier>` | yes | Set a segment's playback speed. |
 | `volume` | `capcut volume <project> <id> <level>` | yes | Set a segment's volume (0.0-1.0). |
 | `trim` | `capcut trim <project> <id> <start> <duration>` | yes | Trim a segment to a start/duration window. |
@@ -20,7 +20,7 @@
 | `export-srt` | `capcut export-srt <project> [options]` | no | Export subtitles to SRT or WebVTT on stdout, per line or per word. |
 | `export-ass` | `capcut export-ass <project> [--karaoke] [--out <file.ass>]` | no | Export styled ASS subtitles on stdout or --out, with per-range overrides and --karaoke word timing. |
 | `export-timeline` | `capcut export-timeline <project> [--out <file.otio>] [--captions skip\|markers]` | no | Export video/audio tracks as OpenTimelineIO JSON for NLE handoff (DaVinci Resolve imports .otio natively). |
-| `import-timeline` | `capcut import-timeline <file.otio> (--out <new-project> \| --into <project>)` | yes | Import OpenTimelineIO JSON (the export-timeline schema set) as a new draft (--out) or append it onto an existing one (--into); unsupported OTIO features are reported, never silent. |
+| `import-timeline` | `capcut import-timeline <file.otio> (--out <new-project> \| --into <project>)` | yes | Import OpenTimelineIO JSON, flattening nested Timeline/Stack/Track sequences, as a new draft (--out) or append it onto an existing one (--into); unsupported OTIO features are reported, never silent. |
 | `materials` | `capcut materials <project> [--type <type>]` | no | List material types and counts; filter with --type. |
 | `segment` | `capcut segment <project> <id>` | no | Full detail for one segment and its material. |
 | `material` | `capcut material <project> <id>` | no | Full detail for one material. |
@@ -31,12 +31,13 @@
 | `crop` | `capcut crop <project> <segment-id> [--ratio <r> \| --rect <x,y,w,h> \| --reset]` | yes | Read or set a video/photo segment's source-material crop (--ratio preset, --rect x,y,w,h, or --reset). |
 | `cut` | `capcut cut <project> <start> <end> --out <path>` | yes | Extract a time range into a new standalone draft. |
 | `duplicate` | `capcut duplicate <project> <segment-id> [--track <track-name>] [--new-track]` | yes | Duplicate a segment at its same timeline position onto a track above the source. |
-| `remove` | `capcut remove <project> <segment-id> [--keep-track] [--keep-materials]` | yes | Remove a segment, its emptied track, and the materials that orphans. |
+| `remove` | `capcut remove <project> <segment-id> [--keep-track] [--keep-materials] [--ripple]` | yes | Remove a segment, its emptied track, and the materials that orphans. |
 | `keyframe` | `capcut keyframe <project> <id> <property> <time> <value> [--easing <name>] \| --batch` | yes | Add a keyframe (position/scale/rotation/alpha/volume); single or --batch. |
 | `transition` | `capcut transition <project> <id> <slug> [--duration <time>]` | yes | Add a transition between segments. |
 | `mask` | `capcut mask <project> <id> <slug> [options] \| --off` | yes | Apply a mask (linear/circle/heart/...) with geometry flags, or --off. |
 | `bg-blur` | `capcut bg-blur <project> <id> <level> \| --off` | yes | Set background blur level 1-4, or --off. |
 | `text-style` | `capcut text-style <project> <id> [options]` | yes | Style text (alpha/shadow/border/background box). |
+| `restyle` | `capcut restyle <project> --preset <preset.json> [--track-name <name>] [options]` | yes | Apply one text-style preset atomically to a whole caption track or every text segment. |
 | `text-anim` | `capcut text-anim <project> <id> [options]` | yes | Add intro/outro/combo text animation. |
 | `image-anim` | `capcut image-anim <project> <id> [options]` | yes | Add intro/outro/combo animation to an image/video segment. |
 | `add-sticker` | `capcut add-sticker <project> <resource-id> <start> <duration> [options]` | yes | Add a sticker on its own track with transform. |
diff --git a/docs/command-reference.zh-CN.md b/docs/command-reference.zh-CN.md
index 69c8e77..f5e8467 100644
--- a/docs/command-reference.zh-CN.md
+++ b/docs/command-reference.zh-CN.md
@@ -8,13 +8,13 @@
 |---|---|:---:|---|
 | `info` | `capcut info <project>` | 否 | 项目概览与素材汇总。 |
 | `version` | `capcut version <project>` | 否 | 检测 CapCut/剪映版本、schema 标志与支持状态。 |
-| `lint` | `capcut lint <project> [options]` | 否 | 基于 schema 的检查（片段重叠、字幕行长、缺失文件、主轨道空隙、外部媒体）；退出码 0/1/2，可直接用于 CI。 |
+| `lint` | `capcut lint <project> [options]` | 否 | 基于 schema 的检查（片段重叠、字幕行长、缺失文件、主轨道空隙、外部媒体）；`--frame-grid` 可检查并修复精确帧边界；退出码 0/1/2，可直接用于 CI。 |
 | `tracks` | `capcut tracks <project>` | 否 | 列出所有轨道。 |
 | `segments` | `capcut segments <project> [--track <type>]` | 否 | 列出片段及其时间信息；可用 --track <type> 按轨道类型过滤。 |
 | `texts` | `capcut texts <project>` | 否 | 列出所有文本/字幕内容。 |
 | `set-text` | `capcut set-text <project> <id> <text>` | 是 | 修改一个文本片段的内容。 |
 | `shift` | `capcut shift <project> <id> <offset>` | 是 | 按偏移量平移单个片段的时间（如 +0.5s）。 |
-| `shift-all` | `capcut shift-all <project> <offset> [--track <type>]` | 是 | 按偏移量平移所有片段（可用 --track 限定某一类轨道）。 |
+| `shift-all` | `capcut shift-all <project> <offset> [--track <type>] [--from <time>]` | 是 | 按偏移量平移片段；可用 --track 限定轨道类型，并用 --from 从安全边界开始。 |
 | `speed` | `capcut speed <project> <id> <multiplier>` | 是 | 设置片段的播放速度。 |
 | `volume` | `capcut volume <project> <id> <level>` | 是 | 设置片段音量（0.0-1.0）。 |
 | `trim` | `capcut trim <project> <id> <start> <duration>` | 是 | 把片段裁剪到指定的起点/时长窗口。 |
@@ -22,7 +22,7 @@
 | `export-srt` | `capcut export-srt <project> [options]` | 否 | 把字幕导出为 SRT 或 WebVTT（输出到 stdout），按行或按词。 |
 | `export-ass` | `capcut export-ass <project> [--karaoke] [--out <file.ass>]` | 否 | 把带样式的 ASS 字幕导出到 stdout 或 --out，支持按区间的样式覆盖；--karaoke 输出逐词时间。 |
 | `export-timeline` | `capcut export-timeline <project> [--out <file.otio>]` | 否 | 把视频/音频轨道导出为 OpenTimelineIO JSON，交接给 NLE（DaVinci Resolve 原生导入 .otio）。 |
-| `import-timeline` | `capcut import-timeline <file.otio> (--out <new-project> \| --into <project>)` | 是 | 导入 OpenTimelineIO JSON（export-timeline 输出的 schema 集合），生成新草稿（--out）或追加到已有草稿（--into）；不支持的 OTIO 特性一律报告，绝不静默丢弃。 |
+| `import-timeline` | `capcut import-timeline <file.otio> (--out <new-project> \| --into <project>)` | 是 | 导入 OpenTimelineIO JSON，把嵌套的 Timeline/Stack/Track 展平，生成新草稿（--out）或追加到已有草稿（--into）；不支持的 OTIO 特性一律报告，绝不静默丢弃。 |
 | `materials` | `capcut materials <project> [--type <type>]` | 否 | 列出素材类型与数量；用 --type 过滤。 |
 | `segment` | `capcut segment <project> <id>` | 否 | 单个片段及其素材的完整详情。 |
 | `material` | `capcut material <project> <id>` | 否 | 单个素材的完整详情。 |
@@ -33,12 +33,13 @@
 | `crop` | `capcut crop <project> <segment-id> [--ratio <r> \| --rect <x,y,w,h> \| --reset]` | 是 | 读取或设置视频/图片片段的源素材裁剪（--ratio 预设、--rect x,y,w,h 或 --reset）。 |
 | `cut` | `capcut cut <project> <start> <end> --out <path>` | 是 | 把一段时间范围提取为一个独立的新草稿。 |
 | `duplicate` | `capcut duplicate <project> <segment-id> [--track <track-name>] [--new-track]` | 是 | 在相同的时间线位置，把片段复制到源轨道上方的轨道。 |
-| `remove` | `capcut remove <project> <segment-id> [--keep-track] [--keep-materials]` | 是 | 删除片段、被它清空的轨道，以及由此孤立的素材。 |
+| `remove` | `capcut remove <project> <segment-id> [--keep-track] [--keep-materials] [--ripple]` | 是 | 删除片段、被它清空的轨道，以及由此孤立的素材；--ripple 会安全地闭合所有轨道上的空档。 |
 | `keyframe` | `capcut keyframe <project> <id> <property> <time> <value> [--easing <name>] \| --batch` | 是 | 添加关键帧（位置/缩放/旋转/透明度/音量）；单个或 --batch 批量。 |
 | `transition` | `capcut transition <project> <id> <slug> [--duration <time>]` | 是 | 在片段之间添加转场。 |
 | `mask` | `capcut mask <project> <id> <slug> [options] \| --off` | 是 | 应用蒙版（线性/圆形/爱心/……），带几何参数，或用 --off 移除。 |
 | `bg-blur` | `capcut bg-blur <project> <id> <level> \| --off` | 是 | 设置背景模糊等级 1-4，或 --off 关闭。 |
 | `text-style` | `capcut text-style <project> <id> [options]` | 是 | 设置文本样式（透明度/阴影/描边/背景框）。 |
+| `restyle` | `capcut restyle <project> --preset <preset.json> [--track-name <name>] [options]` | 是 | 一次将文本样式预设原子化应用到整条字幕轨道或所有文本片段。 |
 | `text-anim` | `capcut text-anim <project> <id> [options]` | 是 | 添加文字入场/出场/组合动画。 |
 | `image-anim` | `capcut image-anim <project> <id> [options]` | 是 | 为图片/视频片段添加入场/出场/组合动画。 |
 | `add-sticker` | `capcut add-sticker <project> <resource-id> <start> <duration> [options]` | 是 | 在独立轨道上添加贴纸并设置变换。 |
@@ -56,7 +57,7 @@
 | `import-srt` | `capcut import-srt <project> <srt-or-> [options]` | 是 | 导入 SRT 文件/stdin，每条字幕生成一个文本片段。 |
 | `import-ass` | `capcut import-ass <project> <ass-or-> [options]` | 是 | 把 ASS/SSA 字幕文件导入为文本片段，并把内联覆盖标签保留为按区间的样式。 |
 | `text-ranges` | `capcut text-ranges <project> <id> --styles <json-or-@file>` | 是 | 为文本片段应用字节级精确的多样式区间。 |
-| `caption` | `capcut caption <project> (--audio <path> \| --from-segment <id>) [options]` | 是 | 用 whisper 转写音频，生成真正的字幕轨道片段。 |
+| `caption` | `capcut caption <project> (--audio <path> \| --from-segment <id>) [options]` | 是 | 用 whisper 转写音频，生成真正的字幕轨道片段；支持音轨选择、脚本匹配门槛和逐词显现。 |
 | `translate` | `capcut translate <project> --to <language> --out <path> [options]` | 是 | 通过 Anthropic API 把草稿克隆为另一种语言。 |
 | `migrate` | `capcut migrate <project> --from <version> --to <version>` | 是 | 在版本边界之间应用已知的 schema 迁移。 |
 | `add-sfx` | `capcut add-sfx <project> <slug> <start> <duration> [options]` | 是 | 在专用轨道上添加音效。 |
@@ -86,6 +87,6 @@
 | `init` | `capcut init <name> [--template <dir>] [--drafts <dir>]` | 是 | 从模板创建一个新的空草稿。 |
 | `quickstart` | `capcut quickstart <name> [--video <f>] [--audio <f>] [--srt <f>] [--drafts <dir>]` | 是 | 一条命令生成第一个草稿：创建 + 添加一个素材 + lint + 打印“在 CapCut 中打开”的步骤。 |
 | `compile` | `capcut compile <spec.json> [--out <draftdir>] [--data <rows.jsonl\|->] [--check \| --plan]` | 是 | 从声明式 JSON spec 构建草稿（describe 的逆操作）。 |
-| `render` | `capcut render <project> [--out <preview.mp4>] [options]` | 否 | 渲染低清 ffmpeg 代理预览（裁剪+变速+音频，--burn-captions）；不是 CapCut 的最终渲染。 |
+| `render` | `capcut render <project> [--out <preview.mp4>] [options]` | 否 | 渲染低清 ffmpeg 代理预览（裁剪+变速+音频，--burn-captions）；支持 --crf / --video-bitrate，大型滤镜图自动走脚本文件；不是 CapCut 的最终渲染。 |
 | `detect-scenes` | `capcut detect-scenes <video> [options]` | 否 | 检测视频中的场景切换切点（ffmpeg scene 滤镜）；输出切点与片段列表，供 compile/cut 使用。 |
 | `detect-silence` | `capcut detect-silence <media> [options]` | 否 | 检测媒体文件中的静音区间（ffmpeg silencedetect）；输出静音与保留片段列表，供 compile/cut 使用。 |
diff --git a/package-lock.json b/package-lock.json
index 40f4590..5e5005b 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -1,12 +1,12 @@
 {
   "name": "capcut-cli",
-  "version": "0.25.0",
+  "version": "0.26.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "capcut-cli",
-      "version": "0.25.0",
+      "version": "0.26.0",
       "license": "MIT",
       "bin": {
         "capcut": "dist/index.js",
diff --git a/package.json b/package.json
index b8f95f3..12ca693 100644
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "capcut-cli",
-  "version": "0.25.0",
+  "version": "0.26.0",
   "description": "Independent, unofficial CLI to create and edit CapCut projects — build drafts from scratch, add video/audio/text, subtitles, timing, speed, volume, templates, cut long-form to shorts. No API needed. Not affiliated with ByteDance.",
   "type": "module",
   "bin": {
diff --git a/python/capcut/__init__.py b/python/capcut/__init__.py
index a15ec06..38fd8cb 100644
--- a/python/capcut/__init__.py
+++ b/python/capcut/__init__.py
@@ -29,7 +29,7 @@
 from dataclasses import dataclass, field
 from typing import Any, Iterable, List, Mapping, Optional, Sequence, Union
 
-__version__ = "0.1.0"
+__version__ = "0.1.1"
 __all__ = [
     "CapcutError",
     "CliNotFound",
diff --git a/python/pyproject.toml b/python/pyproject.toml
index ab457a1..1f7f803 100644
--- a/python/pyproject.toml
+++ b/python/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "capcut"
-version = "0.1.0"
+version = "0.1.1"
 description = "Python client for capcut-cli: create and edit CapCut / JianYing (剪映) drafts from Python. JSON in, JSON out, no server."
 readme = "README.md"
 requires-python = ">=3.9"
diff --git a/src/align.ts b/src/align.ts
index 93e14a8..c16d2f7 100644
--- a/src/align.ts
+++ b/src/align.ts
@@ -56,6 +56,47 @@ export function normalizeToken(token: string): string {
     .replace(/[^\p{L}\p{N}]+/gu, "");
 }
 
+const CJK_CHAR = /[\p{Script=Han}\p{Script=Hiragana}\p{Script=Katakana}]/u;
+const WORD_CHAR = /[\p{L}\p{N}]/u;
+
+/**
+ * Split a Chinese/Japanese line the way speech recognisers timestamp it:
+ * Han/kana characters are individual alignment units, while Latin/digit runs
+ * stay together. Punctuation is kept with the preceding spoken unit so the
+ * author's visible wording survives even though normalizeToken ignores it.
+ */
+export function tokenizeCjkText(text: string): string[] {
+  const tokens: string[] = [];
+  let latin = "";
+  let leadingPunctuation = "";
+  const flushLatin = () => {
+    if (!latin) return;
+    tokens.push(`${leadingPunctuation}${latin}`);
+    latin = "";
+    leadingPunctuation = "";
+  };
+  for (const ch of text) {
+    if (CJK_CHAR.test(ch)) {
+      flushLatin();
+      tokens.push(`${leadingPunctuation}${ch}`);
+      leadingPunctuation = "";
+    } else if (WORD_CHAR.test(ch)) {
+      latin += ch;
+    } else if (/\s/u.test(ch)) {
+      flushLatin();
+    } else if (latin) {
+      flushLatin();
+      tokens[tokens.length - 1] += ch;
+    } else if (tokens.length > 0) {
+      tokens[tokens.length - 1] += ch;
+    } else {
+      leadingPunctuation += ch;
+    }
+  }
+  flushLatin();
+  return tokens.filter((token) => normalizeToken(token).length > 0);
+}
+
 /**
  * Script text → lines → words. Blank lines are dropped; a line is a unit the
  * caller may keep together as one cue. Tokens with no letters or digits
@@ -64,10 +105,10 @@ export function normalizeToken(token: string): string {
 export function tokenizeScript(text: string): string[][] {
   const lines: string[][] = [];
   for (const rawLine of text.split(/\r?\n/)) {
-    const words = rawLine
-      .trim()
-      .split(/\s+/)
-      .filter((w) => w.length > 0 && normalizeToken(w).length > 0);
+    const trimmed = rawLine.trim();
+    const words = CJK_CHAR.test(trimmed)
+      ? tokenizeCjkText(trimmed)
+      : trimmed.split(/\s+/).filter((w) => w.length > 0 && normalizeToken(w).length > 0);
     if (words.length > 0) lines.push(words);
   }
   return lines;
@@ -197,11 +238,27 @@ export function alignScript(
     for (const word of line) flat.push({ word, line: index });
   });
   const scriptNorm = flat.map((w) => normalizeToken(w.word));
-  const recognizedNorm = recognized.map((w) => normalizeToken(w.word));
+  // Whisper commonly emits a Chinese/Japanese "word" as several characters.
+  // When the script uses character units, expand those recogniser tokens and
+  // divide their measured interval evenly so both sides share one token grid.
+  const cjkScript = flat.some((w) => CJK_CHAR.test(w.word));
+  const timedRecognized = cjkScript
+    ? recognized.flatMap((word) => {
+        const pieces = tokenizeCjkText(word.word);
+        if (pieces.length <= 1) return [word];
+        const span = Math.max(pieces.length, word.endUs - word.startUs);
+        return pieces.map((piece, index) => ({
+          word: piece,
+          startUs: Math.round(word.startUs + (span * index) / pieces.length),
+          endUs: Math.round(word.startUs + (span * (index + 1)) / pieces.length),
+        }));
+      })
+    : recognized;
+  const recognizedNorm = timedRecognized.map((w) => normalizeToken(w.word));
   const { pair, exact } = alignTokens(scriptNorm, recognizedNorm);
 
   const words: AlignedWord[] = flat.map((w, i) => {
-    const r = pair[i] >= 0 ? recognized[pair[i]] : null;
+    const r = pair[i] >= 0 ? timedRecognized[pair[i]] : null;
     return {
       word: w.word,
       line: w.line,
@@ -212,7 +269,7 @@ export function alignScript(
   });
 
   // Interpolate the untimed runs between their timed neighbours.
-  const durations = recognized.map((w) => Math.max(1, w.endUs - w.startUs));
+  const durations = timedRecognized.map((w) => Math.max(1, w.endUs - w.startUs));
   const typical = median(durations) || DEFAULT_WORD_DURATION_US;
   let i = 0;
   while (i < words.length) {
@@ -270,11 +327,11 @@ export function alignScript(
   }
   const report: AlignmentReport = {
     script_words: words.length,
-    recognized_words: recognized.length,
+    recognized_words: timedRecognized.length,
     matched,
     substituted,
     inserted,
-    dropped: recognized.length - usedRecognized.size,
+    dropped: timedRecognized.length - usedRecognized.size,
     match_ratio: words.length === 0 ? 0 : Math.round((matched / words.length) * 1000) / 1000,
   };
   return { words, lines, report };
diff --git a/src/caption.ts b/src/caption.ts
index fba99c3..4576428 100644
--- a/src/caption.ts
+++ b/src/caption.ts
@@ -20,6 +20,8 @@ import { storedTextLength } from "./text-offsets.js";
 
 export interface CaptionOptions {
   audio?: string; // path to audio file; if absent, derived from --from-segment
+  audioStream?: number; // zero-based audio stream inside the input container
+  ffmpegCmd?: string; // ffmpeg binary used only when audioStream is selected
   fromSegment?: string; // segment ID of an audio segment in the draft to caption
   whisperCmd?: string; // shell-out command; e.g. "whisper" or "whisper-cli" or "faster-whisper"
   whisperModel?: string; // model name; default "base"
@@ -29,6 +31,8 @@ export interface CaptionOptions {
   preset?: TextStylePreset; // make-preset file whose base style to apply
   whisperEngine?: "auto" | "openai" | "whisper-cpp" | "faster-whisper";
   karaoke?: boolean;
+  wordReveal?: boolean;
+  minScriptMatch?: number;
   maxWords?: number;
   maxChars?: number;
   maxGapMs?: number;
@@ -69,6 +73,8 @@ export interface CaptionResult {
   engine_name?: string;
   words?: number;
   karaoke?: boolean;
+  word_reveal?: boolean;
+  audio_stream?: number;
   keyword_matches?: number;
   color_cycle?: number;
   /** --script alignment quality (only when a script was given). */
@@ -96,8 +102,30 @@ export function captionDraft(draft: Draft, opts: CaptionOptions): CaptionResult
   if (opts.styleRef && opts.preset) {
     throw new Error("--style-ref and --preset are mutually exclusive. Pass one style source.");
   }
-  const audio = resolveAudio(draft, opts);
-  const transcription = runWhisper(audio, opts);
+  if (opts.karaoke && opts.wordReveal) throw new Error("--karaoke and --word-reveal are mutually exclusive.");
+  if (
+    opts.minScriptMatch !== undefined &&
+    (!Number.isFinite(opts.minScriptMatch) || opts.minScriptMatch < 0 || opts.minScriptMatch > 1)
+  ) {
+    throw new Error("--min-script-match must be a number in range 0..1.");
+  }
+  if (opts.minScriptMatch !== undefined && opts.scriptText === undefined) {
+    throw new Error("--min-script-match requires --script <file>.");
+  }
+  if (opts.audioStream !== undefined && (!Number.isInteger(opts.audioStream) || opts.audioStream < 0)) {
+    throw new Error("--audio-stream must be a zero-based non-negative integer.");
+  }
+  const sourceAudio = resolveAudio(draft, opts);
+  const selectedAudio =
+    opts.audioStream === undefined
+      ? undefined
+      : extractAudioStream(sourceAudio, opts.audioStream, opts.ffmpegCmd ?? "ffmpeg");
+  let transcription: TranscriptionResult;
+  try {
+    transcription = runWhisper(selectedAudio?.path ?? sourceAudio, opts);
+  } finally {
+    selectedAudio?.cleanup();
+  }
   const recognizedWords = transcription.words.length > 0 ? transcription.words : wordsFromCues(transcription.cues);
   // The transcript's script decides how words join into a cue (no space inside
   // Chinese or Japanese) and how many of them make one when the caller set no
@@ -118,6 +146,12 @@ export function captionDraft(draft: Draft, opts: CaptionOptions): CaptionResult
     }
     const aligned = alignScript(lines, recognizedWords);
     scriptReport = aligned.report;
+    if (opts.minScriptMatch !== undefined && aligned.report.match_ratio < opts.minScriptMatch) {
+      throw new Error(
+        `Script match ${aligned.report.match_ratio.toFixed(3)} is below --min-script-match ${opts.minScriptMatch}. ` +
+          `No captions were written; check that --script belongs to this audio.`,
+      );
+    }
     cues = opts.karaoke
       ? groupWords(
           aligned.words,
@@ -138,15 +172,16 @@ export function captionDraft(draft: Draft, opts: CaptionOptions): CaptionResult
           ),
         );
   } else {
-    cues = opts.karaoke
-      ? groupWords(
-          recognizedWords,
-          opts.maxWords ?? grouping.karaokeMaxWords,
-          opts.maxChars ?? grouping.karaokeMaxChars,
-          (opts.maxGapMs ?? 500) * 1000,
-          separator,
-        )
-      : transcription.cues;
+    cues =
+      opts.karaoke || opts.wordReveal
+        ? groupWords(
+            recognizedWords,
+            opts.maxWords ?? grouping.karaokeMaxWords,
+            opts.maxChars ?? grouping.karaokeMaxChars,
+            (opts.maxGapMs ?? 500) * 1000,
+            separator,
+          )
+        : transcription.cues;
   }
   if (cues.length === 0) {
     throw new Error("Whisper produced no cues. Check the audio file is not silent and the model name is valid.");
@@ -196,7 +231,19 @@ export function captionDraft(draft: Draft, opts: CaptionOptions): CaptionResult
         baseBold: Boolean(cueStyle.bold ?? false),
         presetRanges,
       });
-    if (opts.karaoke && cue.words && cue.words.length > 0) {
+    if (opts.wordReveal && cue.words && cue.words.length > 0) {
+      for (let wordIndex = 0; wordIndex < cue.words.length; wordIndex++) {
+        const visible = cue.words
+          .slice(0, wordIndex + 1)
+          .map((word) => word.word)
+          .join(separator);
+        const word = cue.words[wordIndex];
+        const next = cue.words[wordIndex + 1];
+        const endUs = next?.startUs ?? cue.endUs;
+        addCaptionSegment(draft, track, visible, word.startUs, Math.max(word.startUs + 1, endUs), cueStyle);
+        created++;
+      }
+    } else if (opts.karaoke && cue.words && cue.words.length > 0) {
       const fullText = cue.words.map((word) => word.word).join(separator);
       let cursor = 0;
       let cueMatches = 0;
@@ -240,12 +287,14 @@ export function captionDraft(draft: Draft, opts: CaptionOptions): CaptionResult
     track_name: trackName,
     first_cue: { start_us: cues[0].startUs, text: cues[0].text },
     last_cue: { start_us: cues[cues.length - 1].startUs, text: cues[cues.length - 1].text },
-    source_audio: audio,
+    source_audio: sourceAudio,
     engine: opts.whisperCmd ? "shell" : "whisper-cli",
     engine_name: transcription.engine,
     caption_script: textScript,
     words: transcription.words.length,
     karaoke: opts.karaoke ?? false,
+    word_reveal: opts.wordReveal || undefined,
+    audio_stream: opts.audioStream,
     // undefined when the flags are off, so JSON output stays byte-identical.
     keyword_matches: highlightWords.length > 0 ? keywordMatches : undefined,
     color_cycle: colorCycle.length > 0 ? colorCycle.length : undefined,
@@ -253,6 +302,36 @@ export function captionDraft(draft: Draft, opts: CaptionOptions): CaptionResult
   };
 }
 
+export function buildAudioStreamInvocation(audio: string, stream: number, output: string): string[] {
+  return ["-y", "-i", audio, "-map", `0:a:${stream}`, "-vn", "-ac", "1", "-ar", "16000", "-c:a", "pcm_s16le", output];
+}
+
+function extractAudioStream(audio: string, stream: number, command: string): { path: string; cleanup: () => void } {
+  const dir = mkdtempSync(join(tmpdir(), "capcut-caption-audio-"));
+  const output = join(dir, "selected.wav");
+  const cleanup = () => {
+    try {
+      rmSync(dir, { recursive: true, force: true });
+    } catch {
+      /* ignore */
+    }
+  };
+  const args = buildAudioStreamInvocation(audio, stream, output);
+  try {
+    const r = spawnSync(command, args, { encoding: "utf-8", timeout: 300_000, maxBuffer: 16 * 1024 * 1024 });
+    if (r.error || r.status !== 0 || !existsSync(output)) {
+      const detail = r.stderr || r.error?.message || `ffmpeg exited ${r.status}`;
+      throw new Error(
+        `Could not extract audio stream ${stream} from ${audio}: ${detail}\n` + `Tried: ${command} ${args.join(" ")}`,
+      );
+    }
+    return { path: output, cleanup };
+  } catch (error) {
+    cleanup();
+    throw error;
+  }
+}
+
 function addCaptionSegment(
   draft: Draft,
   track: Track,
@@ -365,7 +444,7 @@ function runWhisper(audio: string, opts: CaptionOptions): TranscriptionResult {
   const engine = detectEngine(opts);
   const tmpdirPath = mkdtempSync(join(tmpdir(), "capcut-caption-"));
   try {
-    const json = opts.karaoke === true;
+    const json = opts.karaoke === true || opts.wordReveal === true;
     const invocation = buildWhisperInvocation(engine, audio, model, language, tmpdirPath, json);
     const { args, prefix, extension } = invocation;
     const r = spawnSync(cmd, args, { encoding: "utf-8", timeout: 300_000, maxBuffer: 16 * 1024 * 1024 });
diff --git a/src/command-specs.ts b/src/command-specs.ts
index 20ed2fc..4e6acc6 100644
--- a/src/command-specs.ts
+++ b/src/command-specs.ts
@@ -185,7 +185,7 @@ const usages = {
   texts: "capcut texts <project>",
   "set-text": "capcut set-text <project> <id> <text>",
   shift: "capcut shift <project> <id> <offset>",
-  "shift-all": "capcut shift-all <project> <offset> [--track <type>]",
+  "shift-all": "capcut shift-all <project> <offset> [--track <type>] [--from <time>]",
   speed: "capcut speed <project> <id> <multiplier>",
   volume: "capcut volume <project> <id> <level>",
   trim: "capcut trim <project> <id> <start> <duration>",
@@ -204,12 +204,13 @@ const usages = {
   crop: "capcut crop <project> <segment-id> [--ratio <r> | --rect <x,y,w,h> | --reset]",
   cut: "capcut cut <project> <start> <end> --out <path>",
   duplicate: "capcut duplicate <project> <segment-id> [--track <track-name>] [--new-track]",
-  remove: "capcut remove <project> <segment-id> [--keep-track] [--keep-materials]",
+  remove: "capcut remove <project> <segment-id> [--keep-track] [--keep-materials] [--ripple]",
   keyframe: "capcut keyframe <project> <id> <property> <time> <value> [--easing <name>] | --batch",
   transition: "capcut transition <project> <id> <slug> [--duration <time>]",
   mask: "capcut mask <project> <id> <slug> [options] | --off",
   "bg-blur": "capcut bg-blur <project> <id> <level> | --off",
   "text-style": "capcut text-style <project> <id> [options]",
+  restyle: "capcut restyle <project> --preset <preset.json> [--track-name <name>] [options]",
   "text-anim": "capcut text-anim <project> <id> [options]",
   "image-anim": "capcut image-anim <project> <id> [options]",
   "add-sticker": "capcut add-sticker <project> <resource-id> <start> <duration> [options]",
@@ -297,6 +298,12 @@ const optionsByCommand: Record<string, OptionSpec[]> = {
     }),
     option("no_check_paths", ["--no-check-paths"], "boolean", "Skip local media path checks."),
     option("fix", ["--fix"], "boolean", "Mechanically repair fixable issues and write the draft."),
+    option(
+      "frame_grid",
+      ["--frame-grid"],
+      "boolean",
+      "Check target ranges against draft.fps; --fix snaps boundaries and derives duration.",
+    ),
     option("no_probe", ["--no-probe"], "boolean", "Skip ffprobe media checks (VFR / unreadable media)."),
     option(
       "pip",
@@ -308,7 +315,15 @@ const optionsByCommand: Record<string, OptionSpec[]> = {
     FFPROBE,
   ],
   segments: [TRACK],
-  "shift-all": [TRACK],
+  "shift-all": [
+    TRACK,
+    option(
+      "from_time",
+      ["--from"],
+      "time",
+      "Shift only segments starting at or after this boundary; refuses a segment crossing it.",
+    ),
+  ],
   materials: [TRACK],
   "add-audio": [
     option("volume", ["--volume"], "number", "Audio volume.", { default: 1 }),
@@ -370,6 +385,12 @@ const optionsByCommand: Record<string, OptionSpec[]> = {
   remove: [
     option("keep_track", ["--keep-track"], "boolean", "Keep the segment's track even when it becomes empty."),
     option("keep_materials", ["--keep-materials"], "boolean", "Skip the orphan-material sweep (run prune later)."),
+    option(
+      "ripple",
+      ["--ripple"],
+      "boolean",
+      "Close the removed time span across every track; refuses crossing segments.",
+    ),
   ],
   keyframe: [
     option("batch", ["--batch"], "boolean", "Read JSONL keyframes from stdin."),
@@ -405,6 +426,7 @@ const optionsByCommand: Record<string, OptionSpec[]> = {
   ],
   "bg-blur": [option("off", ["--off"], "boolean", "Remove background blur.")],
   "text-style": [...TEXT_STYLE, PRESET],
+  restyle: [TRACK_NAME, ...TEXT_STYLE, PRESET],
   "text-anim": [
     option("intro", ["--intro"], "enum", "Intro animation."),
     option("outro", ["--outro"], "enum", "Outro animation."),
@@ -530,6 +552,13 @@ const optionsByCommand: Record<string, OptionSpec[]> = {
       "Known transcript (plain text). Whisper's word timing is kept, the script's wording is used; each non-empty line is a cue boundary. The result's `script` block reports matched/substituted/inserted words.",
     ),
     option("audio", ["--audio"], "path", "Audio input."),
+    option(
+      "audio_stream",
+      ["--audio-stream"],
+      "number",
+      "Zero-based audio stream to extract from the input container before transcription.",
+    ),
+    option("ffmpeg_cmd", ["--ffmpeg-cmd"], "path", "FFmpeg binary used by --audio-stream."),
     option("from_segment", ["--from-segment"], "id", "Audio segment input."),
     option("whisper_cmd", ["--whisper-cmd"], "path", "Whisper binary."),
     option("whisper_engine", ["--whisper-engine"], "enum", "Whisper CLI dialect.", {
@@ -538,6 +567,18 @@ const optionsByCommand: Record<string, OptionSpec[]> = {
     option("whisper_model", ["--whisper-model"], "string", "Whisper model."),
     option("language", ["--language"], "string", "Language code."),
     option("karaoke", ["--karaoke"], "boolean", "Create word-highlight caption ranges."),
+    option(
+      "word_reveal",
+      ["--word-reveal"],
+      "boolean",
+      "Build each cue progressively, adding one timed word at a time.",
+    ),
+    option(
+      "min_script_match",
+      ["--min-script-match"],
+      "number",
+      "Refuse before writing when caption --script exact-token match ratio is below 0..1.",
+    ),
     option(
       "max_words",
       ["--max-words"],
@@ -765,6 +806,21 @@ const optionsByCommand: Record<string, OptionSpec[]> = {
       "string",
       "Video encoder for the proxy (-c:v; default libx264). Hardware encoders like h264_videotoolbox/h264_nvenc/h264_qsv work when the build carries them; validated against `ffmpeg -encoders` before rendering.",
     ),
+    option(
+      "crf",
+      ["--crf"],
+      "number",
+      "Constant-quality value 0..51 (default 28; mutually exclusive with --video-bitrate).",
+      {
+        default: 28,
+      },
+    ),
+    option(
+      "video_bitrate",
+      ["--video-bitrate"],
+      "string",
+      "Target video bitrate such as 2500k or 4M; switches the proxy from CRF to bitrate mode.",
+    ),
     option("burn_captions", ["--burn-captions"], "boolean", "Burn captions."),
     option(
       "soft_captions",
@@ -855,6 +911,7 @@ optionsByCommand["image-anim"] = optionsByCommand["text-anim"];
 //   --data               -> compile (v0.17 one-draft-per-JSONL-row)
 //   --into               -> import-timeline (v0.17 append target)
 //   --encoder            -> render (v0.20 proxy video encoder)
+//   --crf, --video-bitrate -> render (v0.26 proxy quality controls)
 //   --threshold-db, --min-silence, --pad -> detect-silence (v0.20 silence spans)
 //   --text, --text-file, --tts-cmd -> tts (v0.20 voiceover synthesis)
 //   --nested             -> sync-timelines (v0.21 nested Timelines/ repair)
@@ -868,10 +925,14 @@ optionsByCommand["image-anim"] = optionsByCommand["text-anim"];
 //   --window, --similarity, --min-words -> detect-retakes (v0.22); --json also scopes there
 //   --soft-captions      -> render (v0.22 mov_text subtitle stream)
 //   --like, --from-store -> migrate (v0.23 schema-marker restamp from a donor project)
+//   --word-reveal, --min-script-match, --audio-stream -> caption (v0.26 caption controls)
+//   --from -> shift-all; --ripple -> remove (v0.26 boundary-safe ripple editing)
+//   --frame-grid -> lint (v0.26 exact integer timeline preflight)
 // Everywhere else they fall through to the positional stream verbatim, matching
 // pre-release behaviour where these tokens were unknown and preserved.
 export const RELEASE_SCOPED_FLAGS: ReadonlySet<string> = new Set([
   "--add",
+  "--audio-stream",
   "--apply",
   "--bind",
   "--captions",
@@ -881,8 +942,15 @@ export const RELEASE_SCOPED_FLAGS: ReadonlySet<string> = new Set([
   "--data",
   "--easing",
   "--encoder",
+  "--crf",
   "--format",
   "--from-store",
+  "--word-reveal",
+  "--video-bitrate",
+  "--min-script-match",
+  "--from",
+  "--ripple",
+  "--frame-grid",
   "--full",
   "--granularity",
   "--highlight-words",
@@ -949,6 +1017,7 @@ const mutating = new Set([
   "mask",
   "bg-blur",
   "text-style",
+  "restyle",
   "text-anim",
   "image-anim",
   "add-sticker",
@@ -1034,7 +1103,7 @@ export function buildCommandSpecs(commands: readonly string[], summaries: Record
     if (["add-video", "add-audio", "tts", "compile", "detect-scenes", "detect-silence"].includes(name)) {
       prerequisites.push("ffprobe (optional)");
     }
-    if (name === "caption") prerequisites.push("whisper CLI");
+    if (name === "caption") prerequisites.push("whisper CLI", "ffmpeg (only with --audio-stream)");
     if (name === "tts") prerequisites.push("a local TTS CLI via --tts-cmd");
     if (name === "translate") prerequisites.push("ANTHROPIC_API_KEY or --api-key");
     if (["add-video", "add-audio"].includes(name)) prerequisites.push("network for Wikimedia URLs only");
diff --git a/src/factory.ts b/src/factory.ts
index 0b505c7..bdc9a77 100644
--- a/src/factory.ts
+++ b/src/factory.ts
@@ -2251,6 +2251,10 @@ export function pruneOrphanMaterials(draft: Draft): {
 export interface RemoveSegmentOptions {
   keepTrack?: boolean;
   keepMaterials?: boolean;
+  /** Close the removed time span across every track. Refuses when another
+   * segment crosses that span, because silently slicing it would change
+   * source media; remove/trim the named blocker explicitly first. */
+  ripple?: boolean;
 }
 
 export interface RemoveSegmentResult {
@@ -2263,6 +2267,7 @@ export interface RemoveSegmentResult {
   materialsByType: Record<string, { removed: number; kept: number }>;
   durationBefore: number;
   durationAfter: number;
+  rippleShifted: number;
 }
 
 /**
@@ -2279,6 +2284,28 @@ export function removeSegment(draft: Draft, segId: string, opts: RemoveSegmentOp
   if (!found) throw new Error(`Segment not found: ${segId}`);
   const { track, segment, index } = found;
   const durationBefore = draft.duration;
+  const removedStart = segment.target_timerange.start;
+  const removedEnd = removedStart + segment.target_timerange.duration;
+
+  if (opts.ripple) {
+    const blockers = draft.tracks.flatMap((candidateTrack) =>
+      candidateTrack.segments.filter((candidate) => {
+        if (candidate === segment) return false;
+        const start = candidate.target_timerange.start;
+        const end = start + candidate.target_timerange.duration;
+        return start < removedEnd && end > removedStart;
+      }),
+    );
+    if (blockers.length > 0) {
+      throw new Error(
+        `Ripple delete ${removedStart}us-${removedEnd}us crosses ${blockers.length} other segment(s): ` +
+          `${blockers
+            .slice(0, 5)
+            .map((s) => s.id)
+            .join(", ")}. Remove or trim them first; nothing was changed.`,
+      );
+    }
+  }
 
   track.segments.splice(index, 1);
   let trackRemoved = false;
@@ -2287,6 +2314,19 @@ export function removeSegment(draft: Draft, segId: string, opts: RemoveSegmentOp
     trackRemoved = true;
   }
 
+  let rippleShifted = 0;
+  if (opts.ripple) {
+    const amount = removedEnd - removedStart;
+    for (const candidateTrack of draft.tracks) {
+      for (const candidate of candidateTrack.segments) {
+        if (candidate.target_timerange.start >= removedEnd) {
+          candidate.target_timerange.start -= amount;
+          rippleShifted++;
+        }
+      }
+    }
+  }
+
   let materialsRemoved = 0;
   let materialsByType: Record<string, { removed: number; kept: number }> = {};
   if (!opts.keepMaterials) {
@@ -2314,6 +2354,7 @@ export function removeSegment(draft: Draft, segId: string, opts: RemoveSegmentOp
     materialsByType,
     durationBefore,
     durationAfter: maxEnd,
+    rippleShifted,
   };
 }
 
diff --git a/src/index.ts b/src/index.ts
index 922e606..e058f40 100644
--- a/src/index.ts
+++ b/src/index.ts
@@ -107,6 +107,7 @@ export const COMMANDS = [
   "mask",
   "bg-blur",
   "text-style",
+  "restyle",
   "text-anim",
   "image-anim",
   "add-sticker",
@@ -186,7 +187,9 @@ Overview (start here):
   materials  <project> --type <type>            List items of one material type
   version    <project>                          Detect CapCut/JianYing version + schema flags + support status
   lint       <project>                          Schema-aware checks (overlaps, line length, missing
-             files, main-track gaps, media outside the draft folder)
+             files, main-track gaps, media outside the draft folder).
+             --frame-grid checks every segment boundary against draft fps;
+             combine with --fix to snap start/end together.
              Options:
                --max-chars <n>     Caption line cap (default 42)
                --max-cue-secs <n>  Caption duration cap (default 7)
@@ -266,6 +269,8 @@ Preview:
                                   has no mov_text encoder.
                --ffmpeg-cmd <p>   ffmpeg binary (default ffmpeg)
                --encoder <name>   Video encoder for -c:v (default libx264)
+               --crf <0..51>      Constant quality (default 28)
+               --video-bitrate <r> Target bitrate such as 2500k or 4M
                --dry-run          Print the ffmpeg plan; do not execute
 
 Analyze:
@@ -378,7 +383,9 @@ Add:
 Edit:
   set-text   <project> <id> <text>              Change text content
   shift      <project> <id> <offset>            Shift segment timing (e.g. +0.5s, -1s)
-  shift-all  <project> <offset> [--track <type>] Shift all segments on a track
+  shift-all  <project> <offset> [--track <type>] [--from <time>]
+                                               Shift all segments, optionally
+                                               only from an exact boundary
   speed      <project> <id> <multiplier>        Set playback speed
   volume     <project> <id> <level>             Set volume (0.0-1.0)
   trim       <project> <id> <start> <duration>  Trim segment (times in seconds)
@@ -408,13 +415,15 @@ Edit:
              per-segment companion (speed, canvas, mask, ...) are cloned with
              fresh ids, so edits like crop/mix-mode on the copy never touch
              the source segment.
-  remove     <project> <segment-id> [--keep-track] [--keep-materials]
+  remove     <project> <segment-id> [--keep-track] [--keep-materials] [--ripple]
              Delete a segment in place. A track left empty by the removal is
              dropped too (--keep-track keeps it). Materials no surviving
              segment references are swept in the same pass prune runs —
              including materials that were already orphaned (--keep-materials
              skips the sweep). Recomputes the project duration to the max
-             remaining segment end across all tracks. Undo with restore.
+             remaining segment end across all tracks. --ripple closes the
+             removed span across every track and refuses crossing segments.
+             Undo with restore.
   export-srt <project> [options]                Export subtitles to SRT/WebVTT
   export-ass <project> [--karaoke] [--out <f.ass>]  Export styled subtitles as ASS
   export-timeline <project> [--out <f.otio>] [--captions markers]  Export the cut as OpenTimelineIO for an NLE
@@ -548,6 +557,9 @@ Animate:
                --bg-width --bg-height --bg-h-offset --bg-v-offset
                --preset <file>  Apply a make-preset style preset; explicit
                                 flags override preset values
+  restyle    <project> --preset <file> [--track-name <name>] [options]
+             Apply one preset atomically to every text segment, or only the
+             named caption track. Explicit style flags override the preset.
   text-ranges <project> <id> --styles @path.json  |  --styles '<inline-json>'
              Multi-colour text — write multiple styles to one text segment.
              JSON array of { "start": int, "end": int,
@@ -671,6 +683,8 @@ Caption (v0.4 — real subtitle objects, fixes import-srt mimicry):
              Options:
                --whisper-cmd <cmd>  Path to whisper binary (default: "whisper")
                --whisper-model <m>  Model name (default: "base")
+               --audio-stream <n>   Zero-based audio stream in the input
+               --ffmpeg-cmd <cmd>   ffmpeg used to extract --audio-stream
                --language <code>    ISO code or "auto" (default)
                --track-name <s>     Caption track name (default: "captions")
                --style-ref <seg-id> Mirror styling from existing text segment
@@ -698,6 +712,10 @@ Caption (v0.4 — real subtitle objects, fixes import-srt mimicry):
                                     substituted / inserted words; below 50%
                                     matched it warns that the script probably
                                     belongs to other audio.
+               --min-script-match <0..1>
+                                    Refuse before writing when the script's
+                                    exact-token match ratio is below this floor
+               --word-reveal       Progressive one-word-at-a-time captions
              Precedence: --color-cycle wins over the style-ref/preset base
              colour per cue; keyword emphasis sits on top of base/karaoke
              styling and overrides the matched words' colour/size; with
@@ -866,6 +884,8 @@ interface Flags {
   batch: boolean;
   like?: string;
   fromStore?: boolean;
+  fromTime?: string;
+  ripple?: boolean;
   track?: string;
   out?: string;
   fontSize?: number;
@@ -887,6 +907,8 @@ interface Flags {
   captions?: string;
   // caption --script
   script?: string;
+  wordReveal?: boolean;
+  minScriptMatch?: number;
   // detect-retakes
   window?: number;
   similarity?: number;
@@ -895,6 +917,7 @@ interface Flags {
   softCaptions?: boolean;
   // lint
   pip?: boolean;
+  frameGrid?: boolean;
   // catalogue
   kind?: string;
   // import-srt / import-ass
@@ -983,6 +1006,7 @@ interface Flags {
   fix?: boolean;
   // caption
   audio?: string;
+  audioStream?: number;
   fromSegment?: string;
   whisperCmd?: string;
   whisperEngine?: "auto" | "openai" | "whisper-cpp" | "faster-whisper";
@@ -1040,6 +1064,8 @@ interface Flags {
   fps?: number;
   ffmpegCmd?: string;
   encoder?: string;
+  crf?: number;
+  videoBitrate?: string;
   burnCaptions?: boolean;
   allVideoTracks?: boolean;
   progress?: boolean;
@@ -1197,10 +1223,20 @@ function parseFlags(args: string[]): { positional: string[]; flags: Flags } {
       flags.like = args[++i];
     } else if (a === "--from-store") {
       flags.fromStore = true;
+    } else if (a === "--from" && command === "shift-all" && i + 1 < args.length) {
+      flags.fromTime = args[++i];
+    } else if (a === "--ripple") {
+      flags.ripple = true;
     } else if (a === "--captions" && i + 1 < args.length) {
       flags.captions = args[++i];
     } else if (a === "--script" && i + 1 < args.length) {
       flags.script = args[++i];
+    } else if (a === "--word-reveal") {
+      flags.wordReveal = true;
+    } else if (a === "--min-script-match" && i + 1 < args.length) {
+      flags.minScriptMatch = parseFloat(args[++i]);
+    } else if (a === "--audio-stream" && i + 1 < args.length) {
+      flags.audioStream = Number(args[++i]);
     } else if (a === "--window" && i + 1 < args.length) {
       flags.window = parseFloat(args[++i]);
     } else if (a === "--similarity" && i + 1 < args.length) {
@@ -1445,6 +1481,10 @@ function parseFlags(args: string[]): { positional: string[]; flags: Flags } {
       flags.ffmpegCmd = args[++i];
     } else if (a === "--encoder" && i + 1 < args.length) {
       flags.encoder = args[++i];
+    } else if (a === "--crf" && i + 1 < args.length) {
+      flags.crf = parseFloat(args[++i]);
+    } else if (a === "--video-bitrate" && i + 1 < args.length) {
+      flags.videoBitrate = args[++i];
     } else if (a === "--burn-captions") {
       flags.burnCaptions = true;
     } else if (a === "--all-video-tracks") {
@@ -1469,6 +1509,8 @@ function parseFlags(args: string[]): { positional: string[]; flags: Flags } {
       flags.nested = true;
     } else if (a === "--pip") {
       flags.pip = true;
+    } else if (a === "--frame-grid") {
+      flags.frameGrid = true;
     } else if (a === "--kind" && i + 1 < args.length) {
       flags.kind = args[++i];
     } else if (a === "--clone-style") {
@@ -1800,16 +1842,66 @@ function cmdShift(draft: Draft, filePath: string, segId: string, offsetStr: stri
 
 function cmdShiftAll(draft: Draft, filePath: string, offsetStr: string, flags: Flags, save = true): void {
   const offset = parseTimeInput(offsetStr);
+  const boundary = flags.fromTime === undefined ? null : parseTimeInput(flags.fromTime);
   const tracks = flags.track ? getTracksByType(draft, flags.track) : draft.tracks;
+  if (boundary !== null) {
+    const blocker = tracks
+      .flatMap((track) => track.segments)
+      .find(
+        (seg) =>
+          seg.target_timerange.start < boundary &&
+          seg.target_timerange.start + seg.target_timerange.duration > boundary,
+      );
+    if (blocker) {
+      die(`--from ${flags.fromTime} crosses segment ${blocker.id}; trim or split it first. Nothing was changed.`);
+    }
+    for (const track of tracks) {
+      const beforeBoundary = track.segments.filter((seg) => seg.target_timerange.start < boundary);
+      const moved = track.segments
+        .filter((seg) => seg.target_timerange.start >= boundary)
+        .sort((a, b) => a.target_timerange.start - b.target_timerange.start);
+      const preceding = beforeBoundary.reduce<Segment | null>((latest, seg) => {
+        const end = seg.target_timerange.start + seg.target_timerange.duration;
+        const latestEnd = latest ? latest.target_timerange.start + latest.target_timerange.duration : -1;
+        return end > latestEnd ? seg : latest;
+      }, null);
+      if (preceding && moved.length > 0) {
+        const precedingEnd = preceding.target_timerange.start + preceding.target_timerange.duration;
+        const firstStart = Math.max(0, moved[0].target_timerange.start + offset);
+        if (firstStart < precedingEnd) {
+          die(
+            `--from shift would overlap segments ${preceding.id} and ${moved[0].id} on track "${track.name}". ` +
+              "Use a smaller negative offset or move the boundary; nothing was changed.",
+          );
+        }
+      }
+      for (let index = 1; index < moved.length; index++) {
+        const previous = moved[index - 1];
+        const current = moved[index];
+        const oldPreviousEnd = previous.target_timerange.start + previous.target_timerange.duration;
+        const oldOverlap = oldPreviousEnd > current.target_timerange.start;
+        const newPreviousEnd =
+          Math.max(0, previous.target_timerange.start + offset) + previous.target_timerange.duration;
+        const newCurrentStart = Math.max(0, current.target_timerange.start + offset);
+        if (!oldOverlap && newPreviousEnd > newCurrentStart) {
+          die(
+            `--from shift would overlap segments ${previous.id} and ${current.id} on track "${track.name}". ` +
+              "Use a smaller negative offset; nothing was changed.",
+          );
+        }
+      }
+    }
+  }
   let count = 0;
   for (const track of tracks) {
     for (const seg of track.segments) {
+      if (boundary !== null && seg.target_timerange.start < boundary) continue;
       seg.target_timerange.start = Math.max(0, seg.target_timerange.start + offset);
       count++;
     }
   }
   if (save) saveDraft(filePath, draft);
-  out({ ok: true, shifted: count, offset_us: offset }, flags);
+  out({ ok: true, shifted: count, offset_us: offset, ...(boundary === null ? {} : { from_us: boundary }) }, flags);
 }
 
 function cmdSpeed(draft: Draft, filePath: string, segId: string, multiplier: string, flags: Flags, save = true): void {
@@ -2935,7 +3027,7 @@ async function cmdTextStyle(draft: Draft, filePath: string, positional: string[]
   const applied: string[] = [];
   let materialId = "";
   if (flags.preset) {
-    const presetResult = applyTextPreset(draft, segId, loadPresetFile(flags.preset));
+    const presetResult = applyTextPreset(draft, segId, presetWithFlagOverrides(loadPresetFile(flags.preset), flags));
     materialId = presetResult.materialId;
     applied.push(...presetResult.applied);
   }
@@ -2948,6 +3040,45 @@ async function cmdTextStyle(draft: Draft, filePath: string, positional: string[]
   out({ ok: true, id: segId, material_id: materialId, applied }, flags);
 }
 
+async function cmdRestyle(draft: Draft, filePath: string, flags: Flags): Promise<void> {
+  const { setTextStyle } = await import("./decorators.js");
+  const { applyTextPreset, loadPresetFile } = await import("./preset.js");
+  if (!flags.preset) die("restyle requires --preset <preset.json>.");
+  const tracks = draft.tracks.filter(
+    (track) => track.type === "text" && (flags.trackName === undefined || track.name === flags.trackName),
+  );
+  if (tracks.length === 0) {
+    die(flags.trackName ? `No text track named: ${flags.trackName}` : "Draft has no text tracks.");
+  }
+  const preset = presetWithFlagOverrides(loadPresetFile(flags.preset), flags);
+  const styleOpts = textStyleOptsFromFlags(flags);
+  const segments = tracks.flatMap((track) => track.segments);
+  const applied = new Set<string>();
+  for (const segment of segments) {
+    for (const field of applyTextPreset(draft, segment.id, preset).applied) applied.add(field);
+    for (const field of setTextStyle(draft, segment.id, styleOpts).applied) applied.add(field);
+    if ((flags.x !== undefined || flags.y !== undefined) && segment.clip) {
+      segment.clip.transform = {
+        x: flags.x ?? segment.clip.transform.x,
+        y: flags.y ?? segment.clip.transform.y,
+      };
+      applied.add("transform");
+    }
+  }
+  saveDraft(filePath, draft);
+  out(
+    {
+      ok: true,
+      preset: flags.preset,
+      track_name: flags.trackName ?? null,
+      tracks: tracks.length,
+      segments: segments.length,
+      applied: [...applied],
+    },
+    flags,
+  );
+}
+
 async function cmdTextAnim(draft: Draft, filePath: string, positional: string[], flags: Flags): Promise<void> {
   const { addTextAnim, textAnimSlugs } = await import("./decorators.js");
   const segId = positional[2];
@@ -3225,6 +3356,7 @@ async function cmdRemove(draft: Draft, filePath: string, positional: string[], f
   const result = removeSegment(draft, positional[2], {
     keepTrack: flags.keepTrack,
     keepMaterials: flags.keepMaterials,
+    ripple: flags.ripple,
   });
   saveDraft(filePath, draft);
   out(
@@ -3239,6 +3371,7 @@ async function cmdRemove(draft: Draft, filePath: string, positional: string[], f
       materials_by_type: result.materialsByType,
       duration_before_us: result.durationBefore,
       duration_after_us: result.durationAfter,
+      ripple_shifted: flags.ripple ? result.rippleShifted : undefined,
     },
     flags,
   );
@@ -3694,6 +3827,7 @@ async function cmdLint(draft: Draft, filePath: string, flags: Flags): Promise<{
     ffprobeCmd: flags.ffprobeCmd,
     draftDir: path.dirname(path.resolve(filePath)),
     dryRun: isDryRun(),
+    frameGrid: flags.frameGrid,
   };
   // template-stale (#67, #111) needs the store's newest app version, which
   // costs one timeline read per sibling project — so it is looked up only for
@@ -3815,6 +3949,19 @@ async function cmdCaption(draft: Draft, filePath: string, flags: Flags): Promise
   if (!flags.audio && !flags.fromSegment) {
     die("Missing --audio <path> or --from-segment <id>. One is required.");
   }
+  if (flags.karaoke && flags.wordReveal) die("--karaoke and --word-reveal are mutually exclusive.");
+  if (
+    flags.minScriptMatch !== undefined &&
+    (!Number.isFinite(flags.minScriptMatch) || flags.minScriptMatch < 0 || flags.minScriptMatch > 1)
+  ) {
+    die("--min-script-match must be a number in range 0..1.");
+  }
+  if (flags.minScriptMatch !== undefined && flags.script === undefined) {
+    die("--min-script-match requires --script <file>.");
+  }
+  if (flags.audioStream !== undefined && (!Number.isInteger(flags.audioStream) || flags.audioStream < 0)) {
+    die("--audio-stream must be a zero-based non-negative integer.");
+  }
   const emphasis = await keywordEmphasisFromFlags(flags);
   let scriptText: string | undefined;
   if (flags.script !== undefined) {
@@ -3823,6 +3970,8 @@ async function cmdCaption(draft: Draft, filePath: string, flags: Flags): Promise
   }
   const result = captionDraft(draft, {
     audio: flags.audio,
+    audioStream: flags.audioStream,
+    ffmpegCmd: flags.ffmpegCmd,
     fromSegment: flags.fromSegment,
     scriptText,
     whisperCmd: flags.whisperCmd,
@@ -3833,6 +3982,8 @@ async function cmdCaption(draft: Draft, filePath: string, flags: Flags): Promise
     styleRef: flags.styleRef,
     preset: flags.preset ? loadPresetFile(flags.preset) : undefined,
     karaoke: flags.karaoke,
+    wordReveal: flags.wordReveal,
+    minScriptMatch: flags.minScriptMatch,
     maxWords: flags.maxWords,
     maxChars: flags.maxChars,
     maxGapMs: flags.maxGapMs,
@@ -4960,7 +5111,7 @@ const SUMMARIES: Record<string, string> = {
   texts: "List all text/subtitle content.",
   "set-text": "Change a text segment's content.",
   shift: "Shift one segment's timing by an offset (e.g. +0.5s).",
-  "shift-all": "Shift all segments (optionally on one --track) by an offset.",
+  "shift-all": "Shift segments by an offset, optionally filtering by track and start time.",
   speed: "Set a segment's playback speed.",
   volume: "Set a segment's volume (0.0-1.0).",
   trim: "Trim a segment to a start/duration window.",
@@ -4970,7 +5121,7 @@ const SUMMARIES: Record<string, string> = {
   "export-timeline":
     "Export video/audio tracks as OpenTimelineIO JSON for NLE handoff (DaVinci Resolve imports .otio natively).",
   "import-timeline":
-    "Import OpenTimelineIO JSON (the export-timeline schema set) as a new draft (--out) or append it onto an existing one (--into); unsupported OTIO features are reported, never silent.",
+    "Import OpenTimelineIO JSON, flattening nested Timeline/Stack/Track sequences, as a new draft (--out) or append it onto an existing one (--into); unsupported OTIO features are reported, never silent.",
   "harvest-enums":
     "Learn store resource ids into the per-user catalogue: from one draft, the whole library (--sync), or by hand (--add).",
   materials: "List material types and counts; filter with --type.",
@@ -4989,6 +5140,7 @@ const SUMMARIES: Record<string, string> = {
   mask: "Apply a mask (linear/circle/heart/...) with geometry flags, or --off.",
   "bg-blur": "Set background blur level 1-4, or --off.",
   "text-style": "Style text (alpha/shadow/border/background box).",
+  restyle: "Apply one text-style preset atomically to a whole caption track or every text segment.",
   "text-anim": "Add intro/outro/combo text animation.",
   "image-anim": "Add intro/outro/combo animation to an image/video segment.",
   "add-sticker": "Add a sticker on its own track with transform.",
@@ -5437,12 +5589,23 @@ async function cmdCompileData(specPath: string, flags: Flags): Promise<void> {
 // executing, so the filter graph is inspectable (and the path is ffmpeg-free).
 async function cmdRender(draft: Draft, filePath: string, flags: Flags): Promise<void> {
   const { buildRenderPlan, renderDraft } = await import("./render.js");
+  if (flags.crf !== undefined && flags.videoBitrate !== undefined) {
+    die("--crf and --video-bitrate are mutually exclusive.");
+  }
+  if (flags.crf !== undefined && (!Number.isInteger(flags.crf) || flags.crf < 0 || flags.crf > 51)) {
+    die("--crf must be an integer in range 0..51.");
+  }
+  if (flags.videoBitrate !== undefined && !/^(?:0\.\d*[1-9]\d*|[1-9]\d*(?:\.\d+)?)[kKmM]?$/.test(flags.videoBitrate)) {
+    die("--video-bitrate must be a positive ffmpeg rate such as 2500k or 4M.");
+  }
   const opts = {
     out: flags.out,
     scale: flags.scale,
     fps: flags.fps,
     ffmpegCmd: flags.ffmpegCmd,
     encoder: flags.encoder,
+    crf: flags.crf,
+    videoBitrate: flags.videoBitrate,
     burnCaptions: flags.burnCaptions,
     softCaptions: flags.softCaptions,
     allVideoTracks: flags.allVideoTracks,
@@ -6091,7 +6254,7 @@ async function main(): Promise<void> {
       await cmdDuplicate(draft, filePath, positional, flags);
       break;
     case "remove":
-      requireArgs(positional, 3, "capcut remove <project> <segment-id> [--keep-track] [--keep-materials]");
+      requireArgs(positional, 3, "capcut remove <project> <segment-id> [--keep-track] [--keep-materials] [--ripple]");
       await cmdRemove(draft, filePath, positional, flags);
       break;
     case "keyframe":
@@ -6114,6 +6277,9 @@ async function main(): Promise<void> {
       requireArgs(positional, 3, "capcut text-style <project> <id> [flags]");
       await cmdTextStyle(draft, filePath, positional, flags);
       break;
+    case "restyle":
+      await cmdRestyle(draft, filePath, flags);
+      break;
     case "text-anim":
       requireArgs(positional, 3, "capcut text-anim <project> <id> [--intro <slug>] [--outro <slug>]");
       await cmdTextAnim(draft, filePath, positional, flags);
diff --git a/src/interchange.ts b/src/interchange.ts
index 27075a5..d745c43 100644
--- a/src/interchange.ts
+++ b/src/interchange.ts
@@ -183,9 +183,10 @@ function clipForSegment(draft: Draft, segment: Segment, rate: number, toFrames:
  * Reads the exact schema set the exporter emits (Timeline.1 / Stack.1 /
  * Track.1 / Clip.1 / Gap.1, ExternalReference / MissingReference,
  * LinearTimeWarp) into a flat plan the command applies through the same
- * factory functions `add-video` / `add-audio` use. Everything the CLI cannot
- * represent — unknown track kinds, transitions, nested stacks, other effects,
- * generator references, markers, a non-zero global start — is REPORTED in
+ * factory functions `add-video` / `add-audio` use. Nested Timeline/Stack/Track
+ * sequences are flattened into ordinary tracks with their offsets preserved.
+ * Everything the CLI cannot represent — unknown track kinds, transitions,
+ * other effects, generator references, markers, a non-zero global start — is REPORTED in
  * `skipped`, never silently dropped (the export-timeline house rule).
  *
  * Speed inverts the exporter's documented LinearTimeWarp relationship
@@ -418,17 +419,122 @@ export function otioToImportPlan(doc: unknown): ImportPlan {
     skipped,
   };
 
-  for (const child of (stack as { children: unknown[] }).children) {
-    const schema = schemaOf(child);
-    const track = child as OtioObject;
-    const label = typeof track.name === "string" && track.name ? track.name : schema || "(unnamed)";
-    if (schema !== "Track.1") {
+  const nestedLabel = (parent: string, node: OtioObject, fallback: string): string => {
+    const own = typeof node.name === "string" && node.name ? node.name : fallback;
+    return parent ? `${parent} / ${own}` : own;
+  };
+
+  const declaredDuration = (node: OtioObject, where: string): number | null => {
+    if (node.source_range === null || node.source_range === undefined) return null;
+    return timeRangeUs(node.source_range, rate, `${where} source_range`).durationUs;
+  };
+
+  const compositionDuration = (value: unknown): number => {
+    const node = value as OtioObject;
+    const schema = schemaOf(node);
+    if (schema === "Gap.1") return timeRangeUs(node.source_range, rate, "nested gap source_range").durationUs;
+    if (schema === "Clip.1") {
+      const source = timeRangeUs(node.source_range, rate, "nested clip source_range").durationUs;
+      const warp = (Array.isArray(node.effects) ? node.effects : []).find(
+        (effect) =>
+          schemaOf(effect) === "LinearTimeWarp.1" &&
+          typeof (effect as { time_scalar?: unknown }).time_scalar === "number" &&
+          Number((effect as { time_scalar: number }).time_scalar) > 0,
+      ) as { time_scalar?: number } | undefined;
+      return Math.round(source / (warp?.time_scalar ?? 1));
+    }
+    if (schema === "Timeline.1") return compositionDuration(node.tracks);
+    if (schema === "Track.1") {
+      const rangeDuration = declaredDuration(node, "nested track");
+      if (rangeDuration !== null) return rangeDuration;
+      return (Array.isArray(node.children) ? node.children : []).reduce(
+        (total, child) => total + compositionDuration(child),
+        0,
+      );
+    }
+    if (schema === "Stack.1") {
+      const rangeDuration = declaredDuration(node, "nested stack");
+      if (rangeDuration !== null) return rangeDuration;
+      return (Array.isArray(node.children) ? node.children : []).reduce(
+        (longest, child) => Math.max(longest, compositionDuration(child)),
+        0,
+      );
+    }
+    return 0;
+  };
+
+  const walkStack = (node: OtioObject, startUs: number, parent: string, root = false): number => {
+    const label = root ? parent : nestedLabel(parent, node, "stack");
+    const rangeDuration = root ? null : declaredDuration(node, label);
+    if (rangeDuration !== null) {
       skipped.push({
         track: label,
-        type: schema || "unknown",
-        reason: "unsupported stack child — only Track.1 imports",
+        type: "source_range",
+        reason:
+          "nested stack source_range trimming is not applied; its declared duration still positions following items",
+      });
+    }
+    if (!root) {
+      const nestedCaptions = captionsFromStackMarkers(node.markers, rate, skipped);
+      plan.captions.push(...nestedCaptions.map((caption) => ({ ...caption, startUs: caption.startUs + startUs })));
+    }
+    if (Array.isArray(node.effects) && node.effects.length > 0) {
+      skipped.push({ track: label, type: "effects", reason: "stack-level effects have no CapCut equivalent" });
+    }
+    let durationUs = 0;
+    for (const child of Array.isArray(node.children) ? node.children : []) {
+      const schema = schemaOf(child);
+      if (schema === "Track.1") {
+        durationUs = Math.max(durationUs, walkTrack(child as OtioObject, startUs, label));
+      } else if (schema === "Stack.1") {
+        durationUs = Math.max(durationUs, walkStack(child as OtioObject, startUs, label));
+      } else if (schema === "Timeline.1") {
+        durationUs = Math.max(durationUs, walkTimeline(child as OtioObject, startUs, label));
+      } else {
+        const childLabel = nestedLabel(label, child as OtioObject, schema || "unknown");
+        skipped.push({
+          track: childLabel,
+          type: schema || "unknown",
+          reason: "unsupported stack child — expected Track.1, Stack.1, or Timeline.1",
+        });
+        durationUs = Math.max(durationUs, compositionDuration(child));
+      }
+    }
+    return rangeDuration ?? durationUs;
+  };
+
+  const walkTimeline = (node: OtioObject, startUs: number, parent: string): number => {
+    const label = nestedLabel(parent, node, "timeline");
+    const nestedStart = node.global_start_time;
+    if (
+      nestedStart &&
+      schemaOf(nestedStart) === "RationalTime.1" &&
+      typeof (nestedStart as { value?: unknown }).value === "number" &&
+      (nestedStart as { value: number }).value !== 0
+    ) {
+      skipped.push({ track: label, type: "global_start_time", reason: "non-zero nested timeline start is ignored" });
+    }
+    const tracks = node.tracks;
+    if (schemaOf(tracks) !== "Stack.1") {
+      skipped.push({
+        track: label,
+        type: schemaOf(tracks) || "tracks",
+        reason: "nested timeline has no Stack.1 tracks",
+      });
+      return 0;
+    }
+    return walkStack(tracks as OtioObject, startUs, label);
+  };
+
+  const walkTrack = (track: OtioObject, startUs: number, parent: string): number => {
+    const label = nestedLabel(parent, track, "track");
+    const rangeDuration = declaredDuration(track, label);
+    if (rangeDuration !== null) {
+      skipped.push({
+        track: label,
+        type: "source_range",
+        reason: "track source_range trimming is not applied; its declared duration still positions following items",
       });
-      continue;
     }
     const kind = IMPORTABLE_TRACK_KINDS[String(track.kind)];
     if (!kind) {
@@ -437,7 +543,7 @@ export function otioToImportPlan(doc: unknown): ImportPlan {
         type: String(track.kind ?? "unknown"),
         reason: "unsupported track kind — only Video and Audio tracks import",
       });
-      continue;
+      return rangeDuration ?? compositionDuration(track);
     }
     if (Array.isArray(track.effects) && track.effects.length > 0) {
       skipped.push({ track: label, type: "effects", reason: "track-level effects have no CapCut equivalent" });
@@ -447,7 +553,7 @@ export function otioToImportPlan(doc: unknown): ImportPlan {
     }
 
     const clips: ImportClipPlan[] = [];
-    let cursorUs = 0;
+    let cursorUs = startUs;
     for (const item of Array.isArray(track.children) ? (track.children as unknown[]) : []) {
       const itemSchema = schemaOf(item);
       if (itemSchema === "Gap.1") {
@@ -455,24 +561,41 @@ export function otioToImportPlan(doc: unknown): ImportPlan {
         plan.gaps++;
         continue;
       }
-      if (itemSchema !== "Clip.1") {
-        skipped.push({
-          track: label,
-          type: itemSchema || "unknown",
-          reason:
-            "unsupported timeline item — only Clip.1 and Gap.1 import (transitions overlap, they consume no time)",
-        });
+      if (itemSchema === "Clip.1") {
+        const clip = clipPlan(item as OtioObject, rate, label, skipped);
+        if (!clip) continue;
+        clip.targetStartUs = cursorUs;
+        cursorUs += clip.targetDurationUs;
+        clips.push(clip);
+        plan.clips++;
+        continue;
+      }
+      if (itemSchema === "Stack.1") {
+        cursorUs += walkStack(item as OtioObject, cursorUs, label);
         continue;
       }
-      const clip = clipPlan(item as OtioObject, rate, label, skipped);
-      if (!clip) continue;
-      clip.targetStartUs = cursorUs;
-      cursorUs += clip.targetDurationUs;
-      clips.push(clip);
-      plan.clips++;
+      if (itemSchema === "Track.1") {
+        cursorUs += walkTrack(item as OtioObject, cursorUs, label);
+        continue;
+      }
+      if (itemSchema === "Timeline.1") {
+        cursorUs += walkTimeline(item as OtioObject, cursorUs, label);
+        continue;
+      }
+      skipped.push({
+        track: label,
+        type: itemSchema || "unknown",
+        reason:
+          "unsupported timeline item — clips, gaps, and nested timelines import (transitions overlap, they consume no time)",
+      });
     }
-    plan.tracks.push({ kind, name: typeof track.name === "string" ? track.name : "", clips });
-  }
+    if (clips.length > 0 || (Array.isArray(track.children) && track.children.length === 0)) {
+      plan.tracks.push({ kind, name: label, clips });
+    }
+    return rangeDuration ?? cursorUs - startUs;
+  };
+
+  walkStack(stack as OtioObject, 0, "", true);
 
   return plan;
 }
diff --git a/src/lint.ts b/src/lint.ts
index 09fa848..ec7d1f9 100644
--- a/src/lint.ts
+++ b/src/lint.ts
@@ -10,6 +10,7 @@ import { ffprobeAvailable, isVfr, probeMedia } from "./probe.js";
 import { type CaptionScript, CJK_SCRIPT_LIMITS, captionScript, type ScriptLimit, type ScriptLimits } from "./script.js";
 import { assessMediaRegistrationAt } from "./store.js";
 import { rangesLookDoubled, repairDoubledRanges } from "./text-offsets.js";
+import { quantizeToFrame } from "./time.js";
 import { allUserEnumIds } from "./user-enums.js";
 import { atLeast } from "./version.js";
 
@@ -60,6 +61,7 @@ const FIXABLE_CODES = new Set<string>([
   "media-outside-draft",
   "media-unlinked",
   "text-range-doubled",
+  "segment-off-frame-grid",
 ]);
 
 // Floor for any duration --fix writes: 100ms = three frames at the 30fps
@@ -150,6 +152,11 @@ export interface LintOptions {
   /** Script-specific replacements for maxCharsPerLine / maxCharsPerSecond
    * (see CJK_SCRIPT_LIMITS). Null applies the Latin values to every script. */
   scriptLimits?: ScriptLimits | null;
+  /** Validate every target start/end against draft.fps. Opt-in because older
+   * app-authored drafts commonly carry millisecond timestamps. With --fix,
+   * start and end are snapped together so duration is derived, never rounded
+   * independently into a 1us overlap. */
+  frameGrid?: boolean;
 }
 
 export const DEFAULT_LINT_OPTIONS: LintOptions = {
@@ -174,6 +181,28 @@ export function lintDraft(draft: Draft, opts: LintOptions = DEFAULT_LINT_OPTIONS
     return materialIds.has(id);
   };
 
+  if (opts.frameGrid) {
+    const fps = typeof draft.fps === "number" && Number.isFinite(draft.fps) && draft.fps > 0 ? draft.fps : 30;
+    for (const track of draft.tracks) {
+      for (const segment of track.segments) {
+        const start = segment.target_timerange.start;
+        const end = start + segment.target_timerange.duration;
+        const snappedStart = quantizeToFrame(start, fps);
+        const snappedEnd = quantizeToFrame(end, fps);
+        if (start === snappedStart && end === snappedEnd) continue;
+        issues.push({
+          severity: "warning",
+          code: "segment-off-frame-grid",
+          message:
+            `Segment ${shortId(segment.id)} is off the ${fps}fps frame grid: ` +
+            `${start}us-${end}us -> ${snappedStart}us-${snappedEnd}us; independently rounded ranges can create 1us overlaps`,
+          fixable: snappedEnd > snappedStart,
+          location: { track: track.name, segment_id: segment.id },
+        });
+      }
+    }
+  }
+
   for (const seg of allSegments(draft)) {
     const s = seg.segment;
     if (!draft.materials || !s.material_id) continue;
@@ -1027,6 +1056,39 @@ export function fixDraft(draft: Draft, opts: LintOptions = DEFAULT_LINT_OPTIONS)
   const before = lintDraft(draft, opts);
   const fixed: LintIssue[] = [];
 
+  // Pass -2: snap the two boundaries, then derive duration. Rounding start
+  // and duration independently is the exact 1us-overlap failure this fixes.
+  if (opts.frameGrid) {
+    const fps = typeof draft.fps === "number" && Number.isFinite(draft.fps) && draft.fps > 0 ? draft.fps : 30;
+    const timelineEnd = (): number => {
+      let maxEnd = 0;
+      for (const track of draft.tracks) {
+        for (const segment of track.segments) {
+          maxEnd = Math.max(maxEnd, segment.target_timerange.start + segment.target_timerange.duration);
+        }
+      }
+      return maxEnd;
+    };
+    const oldMaxEnd = timelineEnd();
+    for (const track of draft.tracks) {
+      for (const segment of track.segments) {
+        const oldStart = segment.target_timerange.start;
+        const oldDuration = segment.target_timerange.duration;
+        const newStart = quantizeToFrame(oldStart, fps);
+        const newEnd = quantizeToFrame(oldStart + oldDuration, fps);
+        if (newEnd <= newStart) continue;
+        segment.target_timerange.start = newStart;
+        segment.target_timerange.duration = newEnd - newStart;
+        if (segment.source_timerange?.duration === oldDuration && (segment.speed ?? 1) === 1) {
+          segment.source_timerange.duration = newEnd - newStart;
+        }
+      }
+    }
+    if (draft.duration === oldMaxEnd) {
+      draft.duration = timelineEnd();
+    }
+  }
+
   // Pass -1: link timeline materials to their draft_materials entries
   // (media-unlinked). Pure JSON on the draft — the sidecar is only read — so
   // it runs under dry-run too and the caller's dry-run save discards it.
diff --git a/src/render.ts b/src/render.ts
index a6b26d1..7e32947 100644
--- a/src/render.ts
+++ b/src/render.ts
@@ -1,5 +1,5 @@
 import { spawnSync } from "node:child_process";
-import { existsSync, writeFileSync } from "node:fs";
+import { existsSync, unlinkSync, writeFileSync } from "node:fs";
 import { dirname, extname, join } from "node:path";
 import type { Draft, Segment } from "./draft.js";
 import { extractText } from "./draft.js";
@@ -31,6 +31,8 @@ export interface RenderOptions {
   fps?: number; // output fps override (default draft.fps or 30)
   ffmpegCmd?: string; // ffmpeg binary (default "ffmpeg")
   encoder?: string; // -c:v video encoder (default libx264; e.g. h264_videotoolbox/h264_nvenc/h264_qsv)
+  crf?: number; // constant-quality value; default 28, mutually exclusive with videoBitrate
+  videoBitrate?: string; // target video bitrate such as 4M; switches from CRF to bitrate mode
   burnCaptions?: boolean; // draw text-track segments onto the video
   softCaptions?: boolean; // mux the text-track cues as a mov_text subtitle stream (see softCaptionsFor)
   allVideoTracks?: boolean; // composite overlay video tracks
@@ -79,6 +81,15 @@ const RENDER_TIMEOUT_MS = 600_000;
  */
 const RENDER_MAX_BUFFER = 64 * 1024 * 1024;
 
+// Keep ordinary plans easy to inspect while moving caption-heavy graphs off
+// the command line well before Windows' practical argv ceiling.
+export const FILTER_SCRIPT_THRESHOLD = 8_192;
+const VIDEO_BITRATE = /^(?:0\.\d*[1-9]\d*|[1-9]\d*(?:\.\d+)?)[kKmM]?$/;
+
+export function filterScriptPathFor(output: string): string {
+  return `${output}.filtergraph.txt`;
+}
+
 const FFMPEG_FAILURE_HINTS: Array<{ pattern: RegExp; explain: (match: RegExpMatchArray) => string }> = [
   {
     pattern: /Unknown decoder '([^']+)'|Decoder \(codec ([^)]+)\) not found/,
@@ -146,6 +157,9 @@ export interface RenderPlan {
   capabilities?: FfmpegCapabilities;
   /** Present only with --soft-captions and at least one cue. */
   softCaptions?: SoftCaptionsPlan;
+  /** Long graphs are supplied through a file so Windows command-line limits cannot truncate them. */
+  filterScript?: { path: string; content: string };
+  quality: { mode: "crf"; crf: number } | { mode: "bitrate"; bitrate: string };
 }
 
 export interface RenderResult extends RenderPlan {
@@ -418,6 +432,15 @@ function escapeDrawtext(s: string): string {
  * draft + options (no uuids, no clock) so it can be asserted in tests.
  */
 export function buildRenderPlan(draft: Draft, opts: RenderOptions): RenderPlan {
+  if (opts.crf !== undefined && opts.videoBitrate !== undefined) {
+    throw new Error("render: crf and videoBitrate are mutually exclusive");
+  }
+  if (opts.crf !== undefined && (!Number.isInteger(opts.crf) || opts.crf < 0 || opts.crf > 51)) {
+    throw new Error("render: crf must be an integer in range 0..51");
+  }
+  if (opts.videoBitrate !== undefined && !VIDEO_BITRATE.test(opts.videoBitrate)) {
+    throw new Error("render: videoBitrate must be a positive ffmpeg rate such as 2500k or 4M");
+  }
   const scale = opts.scale && opts.scale > 0 ? opts.scale : 0.5;
   const canvas = draft.canvas_config ?? { width: 1920, height: 1080, ratio: "16:9" };
   const width = Math.max(2, Math.round((canvas.width * scale) / 2) * 2);
@@ -630,11 +653,18 @@ export function buildRenderPlan(draft: Draft, opts: RenderOptions): RenderPlan {
   }
 
   const filterComplex = filterParts.join(";");
+  const filterScript =
+    filterComplex.length > FILTER_SCRIPT_THRESHOLD
+      ? { path: filterScriptPathFor(output), content: filterComplex }
+      : undefined;
+  const quality =
+    opts.videoBitrate !== undefined
+      ? ({ mode: "bitrate", bitrate: opts.videoBitrate } as const)
+      : ({ mode: "crf", crf: opts.crf ?? 28 } as const);
   const args = [
     "-y",
     ...inputArgs,
-    "-filter_complex",
-    filterComplex,
+    ...(filterScript ? ["-filter_complex_script", filterScript.path] : ["-filter_complex", filterComplex]),
     "-map",
     `[${vOut}]`,
     ...(aOut ? ["-map", `[${aOut}]`] : []),
@@ -643,8 +673,7 @@ export function buildRenderPlan(draft: Draft, opts: RenderOptions): RenderPlan {
     opts.encoder ?? "libx264",
     "-preset",
     "veryfast",
-    "-crf",
-    "28",
+    ...(quality.mode === "bitrate" ? ["-b:v", quality.bitrate] : ["-crf", String(quality.crf)]),
     "-pix_fmt",
     "yuv420p",
     ...(aOut ? ["-c:a", "aac", "-b:a", "128k"] : ["-an"]),
@@ -666,6 +695,8 @@ export function buildRenderPlan(draft: Draft, opts: RenderOptions): RenderPlan {
     overlaySegments,
     skipped,
     ...(softCaptions ? { softCaptions } : {}),
+    ...(filterScript ? { filterScript } : {}),
+    quality,
   };
 }
 
@@ -756,6 +787,20 @@ export function renderDraft(draft: Draft, filePath: string, opts: RenderOptions)
   // The SRT must exist before ffmpeg opens its inputs; it stays next to the
   // output afterwards (players auto-load a same-named .srt).
   if (plan.softCaptions) writeFileSync(plan.softCaptions.path, plan.softCaptions.srt, "utf-8");
+  if (plan.filterScript) {
+    try {
+      writeFileSync(plan.filterScript.path, plan.filterScript.content, { encoding: "utf-8", flag: "wx" });
+    } catch (error) {
+      const code = (error as NodeJS.ErrnoException).code;
+      if (code === "EEXIST") {
+        throw new Error(
+          `render: temporary filter script already exists at ${plan.filterScript.path}; ` +
+            "remove or rename that file, or choose another --out path",
+        );
+      }
+      throw error;
+    }
+  }
 
   const cmd = opts.ffmpegCmd ?? "ffmpeg";
   // --progress hands ffmpeg's stderr straight to the terminal. That is both the
@@ -777,6 +822,14 @@ export function renderDraft(draft: Draft, filePath: string, opts: RenderOptions)
       `render: ffmpeg is unavailable at '${cmd}'. Install ffmpeg (\`brew install ffmpeg\` / \`apt install ffmpeg\`) ` +
         `or pass --ffmpeg-cmd <path>. (${e instanceof Error ? e.message : String(e)})`,
     );
+  } finally {
+    if (plan.filterScript) {
+      try {
+        unlinkSync(plan.filterScript.path);
+      } catch {
+        /* ignore cleanup errors; ffmpeg's result remains the primary outcome */
+      }
+    }
   }
   if (r.error) {
     // spawnSync reports its own limits via r.error, not by throwing — a tripped
diff --git a/test/caption-script.test.mjs b/test/caption-script.test.mjs
index 0bb7346..e4ab7a9 100644
--- a/test/caption-script.test.mjs
+++ b/test/caption-script.test.mjs
@@ -4,7 +4,7 @@ import { tmpdir } from "node:os";
 import { dirname, join } from "node:path";
 import { describe, it } from "node:test";
 import { fileURLToPath } from "node:url";
-import { alignScript, normalizeToken, tokenizeScript } from "../dist/align.js";
+import { alignScript, normalizeToken, tokenizeCjkText, tokenizeScript } from "../dist/align.js";
 import { extractText } from "../dist/draft.js";
 import { spawnCli } from "./helpers/spawn-cli.mjs";
 import { tmpDraft } from "./helpers/tmp-draft.mjs";
@@ -128,6 +128,24 @@ describe("align.ts — transcript alignment", () => {
     assert.equal(report.dropped, 2);
     assert.equal(report.match_ratio, 1);
   });
+
+  it("aligns Chinese/Japanese text on characters even when whisper returns a multi-character token", () => {
+    assert.deepEqual(tokenizeCjkText("今天，世界！"), ["今", "天，", "世", "界！"]);
+    const { words, report } = alignScript(tokenizeScript("今天，世界！"), [
+      { word: "今天世界", startUs: 0, endUs: 800_000 },
+    ]);
+    assert.deepEqual(
+      words.map((word) => [word.word, word.startUs, word.endUs]),
+      [
+        ["今", 0, 200_000],
+        ["天，", 200_000, 400_000],
+        ["世", 400_000, 600_000],
+        ["界！", 600_000, 800_000],
+      ],
+    );
+    assert.equal(report.matched, 4);
+    assert.equal(report.match_ratio, 1);
+  });
 });
 
 describe("capcut caption --script (fake whisper)", { skip: isWindows }, () => {
@@ -274,6 +292,38 @@ describe("capcut caption --script (fake whisper)", { skip: isWindows }, () => {
     assert.equal(r.json.script, undefined);
     assert.equal(r.json.cues, 1);
   });
+
+  it("--min-script-match refuses before changing the draft", (t) => {
+    const fix = tmpDraft();
+    t.after(() => fix.cleanup());
+    const s = scratch();
+    t.after(s.cleanup);
+    const audio = join(s.dir, "voice.wav");
+    const heard = join(s.dir, "heard.srt");
+    const script = join(s.dir, "wrong.txt");
+    writeFileSync(audio, "stub");
+    writeFileSync(heard, "1\n00:00:00,000 --> 00:00:01,000\nhello world\n");
+    writeFileSync(script, "quarterly revenue grew quickly\n");
+    const before = readFileSync(fix.path, "utf-8");
+    const r = spawnCli(
+      [
+        "caption",
+        fix.path,
+        "--audio",
+        audio,
+        "--whisper-cmd",
+        FAKE_WHISPER,
+        "--script",
+        script,
+        "--min-script-match",
+        "0.75",
+      ],
+      { env: { FAKE_WHISPER_SOURCE: heard } },
+    );
+    assert.equal(r.status, 1);
+    assert.match(r.stderr, /below --min-script-match 0\.75/);
+    assert.equal(readFileSync(fix.path, "utf-8"), before);
+  });
 });
 
 describe("caption --script surface", () => {
diff --git a/test/caption-stream-reveal.test.mjs b/test/caption-stream-reveal.test.mjs
new file mode 100644
index 0000000..727ed91
--- /dev/null
+++ b/test/caption-stream-reveal.test.mjs
@@ -0,0 +1,130 @@
+import assert from "node:assert/strict";
+import { chmodSync, mkdtempSync, readFileSync, rmSync, writeFileSync } from "node:fs";
+import { tmpdir } from "node:os";
+import { dirname, join } from "node:path";
+import { describe, it } from "node:test";
+import { fileURLToPath } from "node:url";
+import { buildAudioStreamInvocation } from "../dist/caption.js";
+import { extractText } from "../dist/draft.js";
+import { spawnCli } from "./helpers/spawn-cli.mjs";
+import { tmpDraft } from "./helpers/tmp-draft.mjs";
+
+const __dirname = dirname(fileURLToPath(import.meta.url));
+const FAKE_WHISPER = join(__dirname, "helpers", "fake-whisper.mjs");
+const isWindows = process.platform === "win32";
+
+function scratch() {
+  const dir = mkdtempSync(join(tmpdir(), "capcut-caption-stream-"));
+  return { dir, cleanup: () => rmSync(dir, { recursive: true, force: true }) };
+}
+
+function trackTexts(path, name) {
+  const draft = JSON.parse(readFileSync(path, "utf-8"));
+  const track = draft.tracks.find((candidate) => candidate.type === "text" && candidate.name === name);
+  return track.segments
+    .map((segment) => {
+      const material = draft.materials.texts.find((candidate) => candidate.id === segment.material_id);
+      return { text: extractText(material.content), start: segment.target_timerange.start };
+    })
+    .sort((a, b) => a.start - b.start);
+}
+
+describe("caption audio stream selection", () => {
+  it("builds an explicit zero-based ffmpeg stream map", () => {
+    assert.deepEqual(buildAudioStreamInvocation("talk.mov", 2, "selected.wav"), [
+      "-y",
+      "-i",
+      "talk.mov",
+      "-map",
+      "0:a:2",
+      "-vn",
+      "-ac",
+      "1",
+      "-ar",
+      "16000",
+      "-c:a",
+      "pcm_s16le",
+      "selected.wav",
+    ]);
+  });
+
+  it("extracts the selected stream before whisper and reports the original source", { skip: isWindows }, (t) => {
+    const fix = tmpDraft();
+    const s = scratch();
+    t.after(() => fix.cleanup());
+    t.after(s.cleanup);
+    const media = join(s.dir, "multi-audio.mov");
+    const transcript = join(s.dir, "heard.srt");
+    const ffmpeg = join(s.dir, "fake-ffmpeg");
+    writeFileSync(media, "container");
+    writeFileSync(transcript, "1\n00:00:00,000 --> 00:00:01,000\nselected voice\n");
+    writeFileSync(ffmpeg, '#!/bin/sh\nfor last do :; done\nprintf selected > "$last"\n');
+    chmodSync(ffmpeg, 0o755);
+
+    const r = spawnCli(
+      [
+        "caption",
+        fix.path,
+        "--audio",
+        media,
+        "--audio-stream",
+        "1",
+        "--ffmpeg-cmd",
+        ffmpeg,
+        "--whisper-cmd",
+        FAKE_WHISPER,
+        "--track-name",
+        "selected",
+      ],
+      { env: { FAKE_WHISPER_SOURCE: transcript } },
+    );
+    assert.equal(r.status, 0, `stderr: ${r.stderr}`);
+    assert.equal(r.json.audio_stream, 1);
+    assert.equal(r.json.source_audio, media);
+    assert.deepEqual(
+      trackTexts(fix.path, "selected").map((cue) => cue.text),
+      ["selected voice"],
+    );
+  });
+});
+
+describe("caption --word-reveal", { skip: isWindows }, () => {
+  it("creates progressive prefix captions on word timestamps", (t) => {
+    const fix = tmpDraft();
+    const s = scratch();
+    t.after(() => fix.cleanup());
+    t.after(s.cleanup);
+    const audio = join(s.dir, "voice.wav");
+    const transcript = join(s.dir, "heard.json");
+    writeFileSync(audio, "stub");
+    writeFileSync(
+      transcript,
+      JSON.stringify({
+        segments: [
+          {
+            start: 0,
+            end: 1,
+            text: "one two three",
+            words: [
+              { word: "one", start: 0, end: 0.25 },
+              { word: "two", start: 0.3, end: 0.55 },
+              { word: "three", start: 0.6, end: 1 },
+            ],
+          },
+        ],
+      }),
+    );
+    const r = spawnCli(
+      ["caption", fix.path, "--audio", audio, "--whisper-cmd", FAKE_WHISPER, "--word-reveal", "--track-name", "reveal"],
+      { env: { FAKE_WHISPER_SOURCE: transcript } },
+    );
+    assert.equal(r.status, 0, `stderr: ${r.stderr}`);
+    assert.equal(r.json.word_reveal, true);
+    assert.equal(r.json.cues, 3);
+    assert.deepEqual(trackTexts(fix.path, "reveal"), [
+      { text: "one", start: 0 },
+      { text: "one two", start: 300_000 },
+      { text: "one two three", start: 600_000 },
+    ]);
+  });
+});
diff --git a/test/caption.test.mjs b/test/caption.test.mjs
index 1646122..e1a46f7 100644
--- a/test/caption.test.mjs
+++ b/test/caption.test.mjs
@@ -29,6 +29,13 @@ describe("capcut caption — input validation", () => {
     assert.notEqual(r.status, 0);
     assert.match(r.stderr, /Segment not found/);
   });
+
+  it("rejects a non-integer --audio-stream without touching ffmpeg", () => {
+    const r = spawnCli(["caption", fix.path, "--audio", "/nonexistent/audio.mov", "--audio-stream", "1x"]);
+    assert.notEqual(r.status, 0);
+    assert.match(r.stderr, /zero-based non-negative integer/);
+    assert.doesNotMatch(r.stderr, /Audio file not found/);
+  });
 });
 
 describe("caption engine adapters and karaoke grouping", () => {
diff --git a/test/edit.test.mjs b/test/edit.test.mjs
index 9c6622f..205e872 100644
--- a/test/edit.test.mjs
+++ b/test/edit.test.mjs
@@ -120,6 +120,50 @@ describe("capcut shift-all", () => {
       assert.equal(after[i].start_us, oldStarts[i] + 500_000);
     }
   });
+
+  it("--from shifts only segments at or after an exact boundary", () => {
+    const boundary = tmpDraft();
+    try {
+      const before = spawnCli(["segments", boundary.path, "--track", "text"]).json;
+      const r = spawnCli(["shift-all", boundary.path, "+500ms", "--track", "text", "--from", "7.7s"]);
+      assert.equal(r.status, 0, `stderr: ${r.stderr}`);
+      assert.equal(r.json.from_us, 7_700_000);
+      assert.equal(r.json.shifted, 1);
+      const after = spawnCli(["segments", boundary.path, "--track", "text"]).json;
+      assert.deepEqual(
+        after.map((segment) => segment.start_us),
+        before.map((segment) => (segment.start_us >= 7_700_000 ? segment.start_us + 500_000 : segment.start_us)),
+      );
+    } finally {
+      boundary.cleanup();
+    }
+  });
+
+  it("--from refuses to split a segment and leaves the draft unchanged", () => {
+    const boundary = tmpDraft();
+    try {
+      const before = loadDraft(boundary.path);
+      const r = spawnCli(["shift-all", boundary.path, "+500ms", "--track", "text", "--from", "5s"]);
+      assert.equal(r.status, 1);
+      assert.match(r.stderr, /crosses segment/);
+      assert.deepEqual(loadDraft(boundary.path), before);
+    } finally {
+      boundary.cleanup();
+    }
+  });
+
+  it("--from refuses a negative shift that would overlap earlier content", () => {
+    const boundary = tmpDraft();
+    try {
+      const before = loadDraft(boundary.path);
+      const r = spawnCli(["shift-all", boundary.path, "-1s", "--track", "text", "--from", "7.7s"]);
+      assert.equal(r.status, 1);
+      assert.match(r.stderr, /would overlap segments/);
+      assert.deepEqual(loadDraft(boundary.path), before);
+    } finally {
+      boundary.cleanup();
+    }
+  });
 });
 
 describe("capcut speed / volume / opacity", () => {
diff --git a/test/import-timeline.test.mjs b/test/import-timeline.test.mjs
index 4839b66..23a3ec2 100644
--- a/test/import-timeline.test.mjs
+++ b/test/import-timeline.test.mjs
@@ -178,7 +178,6 @@ describe("import-timeline (OTIO)", () => {
       assert.deepEqual(types, [
         "FreezeFrame.1",
         "GeneratorReference.1",
-        "Stack.1",
         "Title",
         "Transition.1",
         "global_start_time",
@@ -190,6 +189,64 @@ describe("import-timeline (OTIO)", () => {
       assert.equal(clip.mediaPath, null);
       assert.equal(clip.speed, 1);
     });
+
+    it("flattens nested Stack/Track sequences while preserving their parent offset", () => {
+      const rt = (value) => ({ OTIO_SCHEMA: "RationalTime.1", rate: 30, value });
+      const range = (start, duration) => ({
+        OTIO_SCHEMA: "TimeRange.1",
+        start_time: rt(start),
+        duration: rt(duration),
+      });
+      const clip = (name, duration) => ({
+        OTIO_SCHEMA: "Clip.1",
+        name,
+        source_range: range(0, duration),
+        effects: [],
+        markers: [],
+        media_reference: { OTIO_SCHEMA: "MissingReference.1", name },
+      });
+      const doc = {
+        OTIO_SCHEMA: "Timeline.1",
+        global_start_time: rt(0),
+        name: "nested-edit",
+        tracks: {
+          OTIO_SCHEMA: "Stack.1",
+          name: "tracks",
+          children: [
+            {
+              OTIO_SCHEMA: "Track.1",
+              kind: "Video",
+              name: "outer",
+              children: [
+                { OTIO_SCHEMA: "Gap.1", source_range: range(0, 30) },
+                {
+                  OTIO_SCHEMA: "Stack.1",
+                  name: "scene",
+                  children: [
+                    {
+                      OTIO_SCHEMA: "Track.1",
+                      kind: "Video",
+                      name: "angle-a",
+                      children: [clip("nested", 60)],
+                    },
+                  ],
+                },
+                clip("after", 30),
+              ],
+            },
+          ],
+        },
+      };
+      const plan = otioToImportPlan(doc);
+      assert.equal(plan.clips, 2);
+      assert.equal(plan.gaps, 1);
+      assert.deepEqual(plan.skipped, []);
+      const nested = plan.tracks.find((track) => track.name === "outer / scene / angle-a");
+      const outer = plan.tracks.find((track) => track.name === "outer");
+      assert.equal(nested.clips[0].targetStartUs, 1_000_000, "nested clip starts after the outer gap");
+      assert.equal(nested.clips[0].targetDurationUs, 2_000_000);
+      assert.equal(outer.clips[0].targetStartUs, 3_000_000, "the following clip starts after the nested sequence");
+    });
   });
 
   describe("round trip (e2e): export-timeline -> import-timeline --out", () => {
diff --git a/test/lint-frame-grid.test.mjs b/test/lint-frame-grid.test.mjs
new file mode 100644
index 0000000..d1b1eaf
--- /dev/null
+++ b/test/lint-frame-grid.test.mjs
@@ -0,0 +1,77 @@
+import assert from "node:assert/strict";
+import { describe, it } from "node:test";
+import { DEFAULT_LINT_OPTIONS, fixDraft, lintDraft } from "../dist/lint.js";
+
+function draftAt(start, duration) {
+  return {
+    fps: 30,
+    duration: start + duration,
+    canvas_config: { width: 1920, height: 1080, ratio: "16:9" },
+    materials: {
+      videos: [{ id: "video", type: "video", path: "", duration: 1_000_000 }],
+      audios: [],
+      texts: [],
+      speeds: [],
+    },
+    tracks: [
+      {
+        id: "track",
+        type: "video",
+        name: "main",
+        attribute: 0,
+        segments: [
+          {
+            id: "segment-off-grid",
+            material_id: "video",
+            target_timerange: { start, duration },
+            source_timerange: { start: 0, duration },
+            speed: 1,
+            extra_material_refs: [],
+          },
+        ],
+      },
+    ],
+  };
+}
+
+const opts = {
+  ...DEFAULT_LINT_OPTIONS,
+  checkLocalPaths: false,
+  probeMedia: false,
+  frameGrid: true,
+};
+
+describe("lint --frame-grid", () => {
+  it("reports off-grid segment boundaries only when explicitly enabled", () => {
+    const draft = draftAt(33_000, 33_667);
+    assert.equal(
+      lintDraft(draft, { ...opts, frameGrid: false }).filter((issue) => issue.code === "segment-off-frame-grid").length,
+      0,
+    );
+    const issues = lintDraft(draft, opts).filter((issue) => issue.code === "segment-off-frame-grid");
+    assert.equal(issues.length, 1);
+    assert.equal(issues[0].fixable, true);
+  });
+
+  it("does not tolerate a one-microsecond boundary drift", () => {
+    const draft = draftAt(33_334, 33_333);
+    assert.equal(lintDraft(draft, opts).filter((issue) => issue.code === "segment-off-frame-grid").length, 1);
+  });
+
+  it("snaps start and end together, then derives duration", () => {
+    const draft = draftAt(33_000, 33_667); // end is already the two-frame boundary
+    const result = fixDraft(draft, opts);
+    const segment = draft.tracks[0].segments[0];
+    assert.equal(segment.target_timerange.start, 33_333);
+    assert.equal(segment.target_timerange.duration, 33_334);
+    assert.equal(segment.source_timerange.duration, 33_334);
+    assert.ok(result.fixed.some((issue) => issue.code === "segment-off-frame-grid"));
+    assert.ok(!result.remaining.some((issue) => issue.code === "segment-off-frame-grid"));
+  });
+
+  it("keeps a root duration that followed the old timeline end in sync", () => {
+    const draft = draftAt(33_000, 33_000);
+    fixDraft(draft, opts);
+    assert.equal(draft.duration, 66_667);
+  });
+});
diff --git a/test/make-preset.test.mjs b/test/make-preset.test.mjs
index 81ba102..3de455d 100644
--- a/test/make-preset.test.mjs
+++ b/test/make-preset.test.mjs
@@ -217,6 +217,48 @@ describe("--preset precedence and validation", () => {
   });
 });
 
+describe("capcut restyle", () => {
+  it("applies one preset to every caption on the selected text track in one command", (t) => {
+    const fix = tmpDraft();
+    const out = tmpDir();
+    t.after(() => fix.cleanup());
+    t.after(out.cleanup);
+    const presetPath = join(out.dir, "bulk.json");
+    writeFileSync(
+      presetPath,
+      JSON.stringify({
+        capcutCliPreset: 1,
+        style: { font_size: 30, text_color: "#FF0000", bold: true },
+      }),
+    );
+
+    const r = spawnCli([
+      "restyle",
+      fix.path,
+      "--preset",
+      presetPath,
+      "--track-name",
+      "Subtitles",
+      "--color",
+      "#00CCFF",
+      "--y",
+      "-0.4",
+    ]);
+    assert.equal(r.status, 0, `stderr: ${r.stderr}`);
+    assert.equal(r.json.tracks, 1);
+    assert.equal(r.json.segments, 3);
+    const draft = loadDraft(fix.path);
+    const track = draft.tracks.find((candidate) => candidate.name === "Subtitles");
+    for (const segment of track.segments) {
+      const material = draft.materials.texts.find((candidate) => candidate.id === segment.material_id);
+      assert.equal(material.text_color, "#00CCFF", "explicit colour wins over the bulk preset");
+      assert.equal(material.font_size, 30);
+      assert.equal(material.bold, true);
+      assert.equal(segment.clip.transform.y, -0.4, "an explicit position works even when the preset has no transform");
+    }
+  });
+});
+
 describe("capcut make-preset errors", () => {
   const fix = tmpDraft();
   const out = tmpDir();
diff --git a/test/remove.test.mjs b/test/remove.test.mjs
index 1dd52b8..bd51797 100644
--- a/test/remove.test.mjs
+++ b/test/remove.test.mjs
@@ -20,6 +20,34 @@ const TXT2 = "cccccc02";
 const TXT3 = "cccccc03";
 
 describe("remove — clean segment removal", () => {
+  it("--ripple closes the removed span across the remaining tracks", (t) => {
+    const fix = tmpDraft();
+    t.after(() => fix.cleanup());
+    const draft = loadDraft(fix.path);
+    draft.tracks = draft.tracks.filter((track) => track.type === "video");
+    draft.materials.audios = [];
+    draft.materials.texts = [];
+    writeFileSync(fix.path, JSON.stringify(draft, null, 2));
+
+    const r = spawnCli(["remove", fix.path, VID1, "--ripple"]);
+    assert.equal(r.status, 0, `stderr: ${r.stderr}`);
+    assert.equal(r.json.ripple_shifted, 1);
+    const remaining = loadDraft(fix.path).tracks[0].segments;
+    assert.equal(remaining.length, 1);
+    assert.equal(remaining[0].target_timerange.start, 0);
+    assert.equal(loadDraft(fix.path).duration, 5_000_000);
+  });
+
+  it("--ripple refuses a span crossed by aligned content and writes nothing", (t) => {
+    const fix = tmpDraft();
+    t.after(() => fix.cleanup());
+    const before = readFileSync(fix.path, "utf-8");
+    const r = spawnCli(["remove", fix.path, VID1, "--ripple"]);
+    assert.equal(r.status, 1);
+    assert.match(r.stderr, /crosses .* other segment/);
+    assert.equal(readFileSync(fix.path, "utf-8"), before);
+  });
+
   it("removes a segment in place, keeps its non-empty track, sweeps its materials", () => {
     const fix = tmpDraft();
     after(() => fix.cleanup());
diff --git a/test/render.test.mjs b/test/render.test.mjs
index cbad03f..3f54196 100644
--- a/test/render.test.mjs
+++ b/test/render.test.mjs
@@ -167,6 +167,51 @@ describe("render — plan (buildRenderPlan, pure)", () => {
     assert.match(on.filterComplex, /drawtext=text='Hook line'/);
   });
 
+  it("supports CRF or target-bitrate quality modes", () => {
+    const s = setup();
+    after(s.cleanup);
+    const crf = buildRenderPlan(buildDraft(s.dir), { out: join(s.dir, "crf.mp4"), crf: 18 });
+    assert.deepEqual(crf.quality, { mode: "crf", crf: 18 });
+    assert.deepEqual(crf.args.slice(crf.args.indexOf("-crf"), crf.args.indexOf("-crf") + 2), ["-crf", "18"]);
+    assert.ok(!crf.args.includes("-b:v"));
+
+    const bitrate = buildRenderPlan(buildDraft(s.dir), { out: join(s.dir, "rate.mp4"), videoBitrate: "4M" });
+    assert.deepEqual(bitrate.quality, { mode: "bitrate", bitrate: "4M" });
+    assert.deepEqual(bitrate.args.slice(bitrate.args.indexOf("-b:v"), bitrate.args.indexOf("-b:v") + 2), [
+      "-b:v",
+      "4M",
+    ]);
+    assert.ok(!bitrate.args.includes("-crf"));
+    assert.throws(
+      () => buildRenderPlan(buildDraft(s.dir), { out: join(s.dir, "bad.mp4"), crf: 18, videoBitrate: "4M" }),
+      /mutually exclusive/,
+    );
+    assert.throws(() => buildRenderPlan(buildDraft(s.dir), { out: join(s.dir, "bad.mp4"), crf: 52 }), /0\.\.51/);
+    assert.throws(
+      () => buildRenderPlan(buildDraft(s.dir), { out: join(s.dir, "bad.mp4"), videoBitrate: "fast" }),
+      /positive ffmpeg rate/,
+    );
+  });
+
+  it("moves a caption-heavy filter graph into a script file argument", () => {
+    const s = setup();
+    after(s.cleanup);
+    const draft = buildDraft(s.dir);
+    const seed = draft.tracks.find((track) => track.type === "text").segments[0];
+    const textTrack = draft.tracks.find((track) => track.type === "text");
+    textTrack.segments = Array.from({ length: 120 }, (_, index) => ({
+      ...seed,
+      id: `caption-${index}`,
+      target_timerange: { start: index * 40_000, duration: 35_000 },
+    }));
+    const plan = buildRenderPlan(draft, { out: join(s.dir, "long.mp4"), burnCaptions: true });
+    assert.ok(plan.filterComplex.length > 8_192);
+    assert.equal(plan.filterScript.content, plan.filterComplex);
+    assert.ok(plan.args.includes("-filter_complex_script"));
+    assert.ok(!plan.args.includes("-filter_complex"));
+    assert.ok(!plan.args.includes(plan.filterComplex), "the long graph must not ride in argv");
+  });
+
   // Regression: `text_color` went into `fontcolor=` raw, so a draft could close
   // the option with a `:` and append drawtext options of its own — `textfile=`
   // reads a local file straight into the burned-in captions.