reelforge 0.6.0 → 0.7.1

package/README.md

@@ -1,237 +1,240 @@
-# reelforge
-
-> CLI for [ReelForge Studio](https://github.com/puke3615/ReelForge) — every REST API exposed as a command, with `--help` available at every level.
-
-## Install
-
-```bash
-npm install -g reelforge
-```
-
-Or use directly without install:
-
-```bash
-npx reelforge <command>
-```
-
-After install, two binaries are on your `PATH` — `reelforge` and the short alias `rf`. Both behave identically; the docs use `rf` from here on.
-
-```bash
-rf --version            # same as `reelforge --version`
-```
-
-## Quick start
-
-The CLI ships pointing at the hosted instance (`https://reelforge.timor419.com`). Log in once, then call:
-
-```bash
-npm install -g reelforge
-rf login                                       # opens browser; headless? rf login <api_key>
-rf whoami                                      # balance + api_keys
-rf create "为什么我们还没找到外星文明？"          # auto-saves to ./<title>-<id>.mp4 in cwd
-```
-
-That's the whole story — no server to run.
-
-### Output behavior
-
-| invocation | result |
-|---|---|
-| `rf create "..."` | Saves to `./<sanitized-title>-<task_id_short>.mp4`, prints the path |
-| `rf create "..." -o ./videos/space.mp4` | Saves to that exact path (must include filename, not just a directory) |
-| `rf create "..." --no-download` | Skips local save, prints JSON result with `video_url` |
-| `rf create "..." \| jq .video_url` | When stdout is piped, download is skipped automatically |
-
-### Self-hosting
-
-If you want to run your own ReelForge Studio (own RelayX key, your own pricing) clone the upstream repo, `pnpm dev`, then point the CLI at it:
-
-```bash
-rf --server http://localhost:8501 health
-# or persist:
-export REELFORGE_SERVER=http://localhost:8501
-# or via `rf login <key> --server http://localhost:8501`
-```
-
-## Global options
-
-| flag | description |
-|---|---|
-| `-s, --server <url>` | ReelForge server URL (overrides `$REELFORGE_SERVER`; default `https://reelforge.timor419.com`) |
-| `-k, --api-key <key>` | API key (overrides `$REELFORGE_API_KEY` and `reelforge login` saved key) |
-| `--json` | Output raw JSON instead of pretty text — pipe-friendly |
-| `--quiet` | Suppress informational messages on stderr |
-| `-v, --version` | Show CLI version |
-| `-h, --help` | Show help (works on every sub-command) |
-
-## Command map
-
-Run `rf <command> --help` for full details on any of these.
-
-### Core capabilities
-
-| command | what it does |
-|---|---|
-| `llm chat -p <text>` | Send one prompt to the configured LLM (RelayX gateway by default) |
-| `llm presets` | List built-in RelayX model presets |
-| `tts edge -t <text> -o out.mp3` | Local Edge TTS synthesis (free) |
-| `tts relayx -t <text> -o out.mp3` | RelayX TTS (vox/index-tts-2, 149 built-in voices) |
-| `tts voices [--locale zh]` | List supported Edge TTS voices |
-| `images generate -p <prompt> -m rx-image-flux` | Image generation via RelayX (rx-image-z / rx-image-flux / rx-image-qwen) |
-
-### Content / audio / subtitle atomics
-
-| command | what it does |
-|---|---|
-| `content scene-plan -t <topic>` | Single LLM call: title + master script + per-scene image prompts (replaces the old narration / split / image-prompts / title trio) |
-| `content scene-plan --script <text-or-@file>` | Same, but the user supplies the script verbatim — LLM only segments and writes image prompts |
-| `audio transcribe -f <file>` / `--url <url>` | RelayX paraformer-v2 ASR with word + segment timestamps |
-| `subtitles split -t <text-or-@file>` | Deterministic tiered-punctuation subtitle line splitter (pure function, zero billing) |
-
-### Composition
-
-| command | what it does |
-|---|---|
-| `templates list [--size 1080x1920] [--type image]` | List HTML frame templates |
-| `templates preview <keyOrPath> [-o out.png]` | Render a preview from a preset key **or your own local .html file** |
-| `templates show <key> [-o file.html]` | Print or save the source HTML of any preset — copy it as a starting point for a custom template |
-| `frames render -t <keyOrPath> --title ... --text ...` | Render a single composed frame to PNG. `-t` accepts a preset key **or a local .html path** |
-| `compositions concat <v1> <v2> -o out.mp4` | FFmpeg concat (+ optional BGM) |
-| `compositions bgm -i video.mp4 --bgm bgm.mp3 -o out.mp4` | Add background music |
-| `compositions image-to-video -i img.png -a aud.mp3 -o out.mp4` | Build video from image + audio |
-| `compositions overlay -v video.mp4 --overlay overlay.png -o out.mp4` | Overlay PNG on video |
-
-### End-to-end pipelines
-
-All `pipelines *` commands submit an **async task** and (by default) poll until it finishes with a live progress indicator on stderr. Use `--no-wait` to return immediately with a `task_id`, then `rf tasks wait <id>` later.
-
-The standard pipeline is **audio-first**: scene-plan → one-shot TTS → ASR alignment → per-scene image generation → per-subtitle-line frame rendering → ffmpeg mux. One continuous master audio track; image cuts at scene boundaries; subtitle cuts at line boundaries.
-
-| command | what it does |
-|---|---|
-| `pipelines standard -t <topic>` (or `--script <text>`) | Audio-first pipeline; `-d/--duration` and `-p/--pace` are the two main knobs |
-
-### Resources
-
-| command | what it does |
-|---|---|
-| `bgm list / upload <file> / delete <name>` | Manage background music |
-| `files list / upload <file> / download <path> / delete <path>` | Manage user assets |
-
-### System
-
-| command | what it does |
-|---|---|
-| `config get` | Read server config (keys masked) |
-| `config set <key> <value>` | Update a dotted-path setting (e.g. `llm.api_key sk-xxx`) |
-| `config patch <file>` | Apply a JSON-merge patch |
-| `tasks list [--status running]` | List recent tasks |
-| `tasks get <id>` / `tasks wait <id>` / `tasks cancel <id>` | Task lifecycle |
-| `history list / get <id> / delete <id>` | Browse / delete completed runs |
-| `health` | Server health + capability check |
-
-## Examples
-
-```bash
-# 1. One-click out a video (45s default, AI writes the script)
-rf create "为什么我们还没找到外星文明？"
-
-# 2. Longer video with a slower visual rhythm
-rf create "深夜便利店的灯光" -d 90 -p slow
-
-# 3. Your own script — no narration-splitting on your side, the pipeline handles it
-rf create --script @./my-script.txt
-rf create --script "雨水缓缓滑落在玻璃窗上，像是无声的泪珠。"
-
-# 4. Pick a built-in visual style preset
-rf create "美食教程" --style photorealistic
-
-# 5. Pipeline form with explicit output path
-rf pipelines standard \
-  --script @./script.txt \
-  --frame-template 1080x1920/image_default.html \
-  -p normal -o smoke.mp4
-
-# 6. Inspect existing tasks & redownload a finished video
-rf tasks list --limit 5
-rf history get <task-id> --download recovered.mp4
-
-# 7. Atomics for stand-alone use
-rf content scene-plan -t "雨天的玻璃窗" -d 45 --json | jq .scenes
-rf audio transcribe -f narration.mp3 --json | jq '.words[:5]'
-rf subtitles split -t @./narration.txt --min 10 --hard-max 24
-
-# 8. JSON pipe for automation
-rf llm presets --json | jq '.[].defaultModel'
-
-# 9. Configure & test LLM (self-hosted)
-rf config set llm.api_key rx-xxxxx          # RelayX key (or your own provider key)
-rf config set llm.base_url https://relayx.timor419.com/v1
-rf config set llm.model anthropic/claude-4-7-sonnet
-rf llm chat -p 'one-sentence summary of antifragile'
-
-# 10. Use your own HTML template (no PR/release needed)
-#     Any --frame-template that points to a local .html file is read and sent
-#     inline. Declare size inside the file via
-#       <meta name="template:width"  content="1080">
-#       <meta name="template:height" content="1920">
-#     or pass --frame-template-size 1080x1920.
-rf templates show 1080x1920/image_default.html -o my-brand.html   # copy a preset
-# ...edit my-brand.html to suit your style...
-rf templates preview ./my-brand.html --title "Hello" -o preview.png
-rf frames render -t ./my-brand.html --values '{"author":"Alice"}' -o frame.png
-rf pipelines standard -t "宠物" --frame-template ./my-brand.html -o final.mp4
-```
-
-### Custom HTML templates
-
-Easiest way to start: grab a preset as a reference.
-
-```bash
-rf templates list                                                  # see all keys
-rf templates show 1080x1920/static_default.html                    # print to stdout
-rf templates show 1080x1920/image_default.html -o my-brand.html    # save and edit
-```
-
-`{{title}}`, `{{text}}`, `{{image}}`, `{{index}}` are reserved built-ins; everything else uses the `{{name:type=default}}` DSL (`type` ∈ `text|number|color|bool`). Pass extras through `--values '{"author":"Alice"}'` (or `template_params` on the pipeline API).
-
-#### Template type — does the pipeline generate an AI image per scene?
-
-When you ship an inline template through `rf create` / `rf pipelines standard`, ReelForge needs to know whether each scene should kick off RelayX image generation. Resolution priority (high → low):
-
-1. Explicit flag — `--frame-template-type image|static|asset` (or `frame_template_type` in the API body).
-2. Inside the HTML — `<meta name="template:type" content="image">` (or `static` / `asset`).
-3. **Default: `image`** — best practice for zero-config users. If your template doesn't reference scene imagery (pure-text card, etc.), declare `static` explicitly to skip image generation and its cost.
-
-The placeholder `{{image}}` no longer doubles as a type signal — declare type explicitly.
-
-Limits and safety:
-
-- Max 2 MB per inline HTML.
-- The render sandbox blocks `file://`, loopback / private / link-local IPs, CGNAT range, cloud-metadata, and `*.local` / `*.internal` hostnames. So your template can only reference public `https`/`http` resources or `data:` URIs.
-- If the CLI is talking to a hosted server, local-path `--image` won't reach the server; either upload to `rf files upload` first or use an HTTPS URL / data: URI.
-
-#### API field reference
-
-| endpoint | inline HTML field | size field | type field |
-|---|---|---|---|
-| `POST /api/v1/frames/render` | `template_html` | `size` | — (n/a, no image generation) |
-| `POST /api/v1/templates/preview` | `template_html` | `size` | — |
-| `POST /api/v1/pipelines/standard` | `frame_template_inline` | `frame_template_size` | `frame_template_type` |
-
-The pipeline endpoint uses the `frame_template_*` prefix because it already has a `frame_template` field (preset key). The single-frame endpoints use the shorter `template_html` because they don't.
-
-## Tip — getting unstuck
-
-Every level has `--help`:
-
-```bash
-rf --help                     # top-level overview
-rf pipelines --help           # list of pipelines
-rf pipelines standard --help  # full option reference
-rf tts edge --help            # one specific command
-```
-
-## License
-
-Apache-2.0
+# reelforge
+
+> CLI for [ReelForge Studio](https://github.com/puke3615/ReelForge) — every REST API exposed as a command, with `--help` available at every level.
+
+## Install
+
+```bash
+npm install -g reelforge
+```
+
+Or use directly without install:
+
+```bash
+npx reelforge <command>
+```
+
+After install, two binaries are on your `PATH` — `reelforge` and the short alias `rf`. Both behave identically; the docs use `rf` from here on.
+
+```bash
+rf --version            # same as `reelforge --version`
+```
+
+## Quick start
+
+The CLI ships pointing at the hosted instance (`https://reelforge.timor419.com`). Log in once, then call:
+
+```bash
+npm install -g reelforge
+rf login                                       # opens browser; headless? rf login <api_key>
+rf whoami                                      # balance + api_keys
+rf create "为什么我们还没找到外星文明？"          # auto-saves to ./<title>-<id>.mp4 in cwd
+```
+
+That's the whole story — no server to run.
+
+### Output behavior
+
+| invocation | result |
+|---|---|
+| `rf create "..."` | Saves to `./<sanitized-title>-<task_id_short>.mp4`, prints the path |
+| `rf create "..." -o ./videos/space.mp4` | Saves to that exact path (must include filename, not just a directory) |
+| `rf create "..." --no-download` | Skips local save, prints JSON result with `video_url` |
+| `rf create "..." \| jq .video_url` | When stdout is piped, download is skipped automatically |
+
+### Self-hosting
+
+If you want to run your own ReelForge Studio (own RelayX key, your own pricing) clone the upstream repo, `pnpm dev`, then point the CLI at it:
+
+```bash
+rf --server http://localhost:8501 health
+# or persist:
+export REELFORGE_SERVER=http://localhost:8501
+# or via `rf login <key> --server http://localhost:8501`
+```
+
+## Global options
+
+| flag | description |
+|---|---|
+| `-s, --server <url>` | ReelForge server URL (overrides `$REELFORGE_SERVER`; default `https://reelforge.timor419.com`) |
+| `-k, --api-key <key>` | API key (overrides `$REELFORGE_API_KEY` and `reelforge login` saved key) |
+| `--json` | Output raw JSON instead of pretty text — pipe-friendly |
+| `--quiet` | Suppress informational messages on stderr |
+| `-v, --version` | Show CLI version |
+| `-h, --help` | Show help (works on every sub-command) |
+
+## Command map
+
+Run `rf <command> --help` for full details on any of these.
+
+### Core capabilities
+
+| command | what it does |
+|---|---|
+| `llm chat -p <text>` | Send one prompt to the configured LLM (RelayX gateway by default) |
+| `llm presets` | List built-in RelayX model presets |
+| `tts edge -t <text> -o out.mp3` | Local Edge TTS synthesis (free) |
+| `tts relayx -t <text> -o out.mp3` | RelayX TTS (vox/index-tts-2, 149 built-in voices) |
+| `tts voices [--locale zh]` | List supported Edge TTS voices |
+| `images generate -p <prompt> -m rx-image-flux` | Image generation via RelayX (rx-image-z / rx-image-flux / rx-image-qwen) |
+
+### Content / audio / subtitle atomics
+
+| command | what it does |
+|---|---|
+| `content scene-plan -t <topic>` | Single LLM call: title + master script + per-scene image prompts (replaces the old narration / split / image-prompts / title trio) |
+| `content scene-plan --script <text-or-@file>` | Same, but the user supplies the script verbatim — LLM only segments and writes image prompts |
+| `audio transcribe -f <file>` / `--url <url>` | RelayX paraformer-v2 ASR with word + segment timestamps |
+| `subtitles split -t <text-or-@file>` | Deterministic tiered-punctuation subtitle line splitter (pure function, zero billing) |
+
+### Composition
+
+| command | what it does |
+|---|---|
+| `templates list [--size 1080x1920] [--type image]` | List HTML frame templates |
+| `templates preview <keyOrPath> [-o out.png]` | Render a preview from a preset key **or your own local .html file** |
+| `templates show <key> [-o file.html]` | Print or save the source HTML of any preset — copy it as a starting point for a custom template |
+| `frames render -t <keyOrPath> --title ... --text ...` | Render a single composed frame to PNG. `-t` accepts a preset key **or a local .html path** |
+| `compositions concat <v1> <v2> -o out.mp4` | FFmpeg concat (+ optional BGM) |
+| `compositions bgm -i video.mp4 --bgm bgm.mp3 -o out.mp4` | Add background music |
+| `compositions image-to-video -i img.png -a aud.mp3 -o out.mp4` | Build video from image + audio |
+| `compositions overlay -v video.mp4 --overlay overlay.png -o out.mp4` | Overlay PNG on video |
+
+### End-to-end pipelines
+
+All `pipelines *` commands submit an **async task** and (by default) poll until it finishes with a live progress indicator on stderr. Use `--no-wait` to return immediately with a `task_id`, then `rf tasks wait <id>` later.
+
+The standard pipeline is **audio-first**: scene-plan → one-shot TTS → ASR alignment → per-scene image generation → per-subtitle-line frame rendering → ffmpeg mux. One continuous master audio track; image cuts at scene boundaries; subtitle cuts at line boundaries.
+
+| command | what it does |
+|---|---|
+| `pipelines standard -t <topic>` (or `--script <text>`) | Audio-first pipeline; `-d/--duration` and `-p/--pace` are the two main knobs |
+
+### Resources
+
+| command | what it does |
+|---|---|
+| `bgm list / upload <file> / delete <name>` | Manage background music |
+| `files list / upload <file> / download <path> / delete <path>` | Manage user assets |
+
+### System
+
+| command | what it does |
+|---|---|
+| `config get` | Read server config (keys masked) |
+| `config set <key> <value>` | Update a dotted-path setting (e.g. `llm.api_key sk-xxx`) |
+| `config patch <file>` | Apply a JSON-merge patch |
+| `tasks list [--status running]` | List recent tasks |
+| `tasks get <id>` / `tasks wait <id>` / `tasks cancel <id>` | Task lifecycle |
+| `history list / get <id> / delete <id>` | Browse / delete completed runs |
+| `health` | Server health + capability check |
+
+## Examples
+
+```bash
+# 1. One-click out a video (45s default, AI writes the script)
+rf create "为什么我们还没找到外星文明？"
+
+# 2. Longer video with a slower visual rhythm
+rf create "深夜便利店的灯光" -d 90 -p slow
+
+# 3. Your own script — no narration-splitting on your side, the pipeline handles it
+rf create --script @./my-script.txt
+rf create --script "雨水缓缓滑落在玻璃窗上，像是无声的泪珠。"
+
+# 4. Pick a built-in visual style preset
+rf create "美食教程" --style photorealistic
+
+# 5. Pipeline form with explicit output path
+rf pipelines standard \
+  --script @./script.txt \
+  --frame-template 1080x1920/image_default.html \
+  -p normal -o smoke.mp4
+
+# 6. Inspect existing tasks & redownload a finished video
+rf tasks list --limit 5
+rf history get <task-id> --download recovered.mp4
+
+# 7. Atomics for stand-alone use
+rf content scene-plan -t "雨天的玻璃窗" -d 45 --json | jq .scenes
+rf audio transcribe -f narration.mp3 --json | jq '.words[:5]'
+rf subtitles split -t @./narration.txt --min 10 --hard-max 24
+
+# 8. JSON pipe for automation
+rf llm presets --json | jq '.[].defaultModel'
+
+# 9. Configure & test LLM (self-hosted)
+rf config set llm.api_key rx-xxxxx          # RelayX key (or your own provider key)
+rf config set llm.base_url https://relayx.timor419.com/v1
+rf config set llm.model anthropic/claude-4-7-sonnet
+rf llm chat -p 'one-sentence summary of antifragile'
+
+# 10. Use your own HTML template (no PR/release needed)
+#     Any --frame-template that points to a local .html file is read and sent
+#     inline. Declare size inside the file via
+#       <meta name="template:width"  content="1080">
+#       <meta name="template:height" content="1920">
+#     or pass --frame-template-size 1080x1920.
+rf templates show 1080x1920/image_default.html -o my-brand.html   # copy a preset
+# ...edit my-brand.html to suit your style...
+rf templates preview ./my-brand.html --title "Hello" -o preview.png
+rf frames render -t ./my-brand.html --values '{"author":"Alice"}' -o frame.png
+rf pipelines standard -t "宠物" --frame-template ./my-brand.html -o final.mp4
+```
+
+### Custom HTML templates
+
+Easiest way to start: grab a preset as a reference.
+
+```bash
+rf templates list                                                  # see all keys
+rf templates show 1080x1920/static_default.html                    # print to stdout
+rf templates show 1080x1920/image_default.html -o my-brand.html    # save and edit
+```
+
+`{{title}}`, `{{text}}`, `{{image}}`, `{{index}}`, `{{total}}` are reserved built-ins auto-injected by the pipeline; everything else uses the `{{name:type=default}}` DSL (`type` ∈ `text|number|color|bool`). Pass extras through `--values '{"author":"Alice"}'` (or `template_params` on the pipeline API).
+
+- `{{index}}` — current scene number, 1-based
+- `{{total}}` — scene count the LLM actually produced (use this for "scene N of M" badges; don't hardcode in `template_params`, the scene count is decided at runtime)
+
+#### Template type — does the pipeline generate an AI image per scene?
+
+When you ship an inline template through `rf create` / `rf pipelines standard`, ReelForge needs to know whether each scene should kick off RelayX image generation. Resolution priority (high → low):
+
+1. Explicit flag — `--frame-template-type image|static|asset` (or `frame_template_type` in the API body).
+2. Inside the HTML — `<meta name="template:type" content="image">` (or `static` / `asset`).
+3. **Default: `image`** — best practice for zero-config users. If your template doesn't reference scene imagery (pure-text card, etc.), declare `static` explicitly to skip image generation and its cost.
+
+The placeholder `{{image}}` no longer doubles as a type signal — declare type explicitly.
+
+Limits and safety:
+
+- Max 2 MB per inline HTML.
+- The render sandbox blocks `file://`, loopback / private / link-local IPs, CGNAT range, cloud-metadata, and `*.local` / `*.internal` hostnames. So your template can only reference public `https`/`http` resources or `data:` URIs.
+- If the CLI is talking to a hosted server, local-path `--image` won't reach the server; either upload to `rf files upload` first or use an HTTPS URL / data: URI.
+
+#### API field reference
+
+| endpoint | inline HTML field | size field | type field |
+|---|---|---|---|
+| `POST /api/v1/frames/render` | `template_html` | `size` | — (n/a, no image generation) |
+| `POST /api/v1/templates/preview` | `template_html` | `size` | — |
+| `POST /api/v1/pipelines/standard` | `frame_template_inline` | `frame_template_size` | `frame_template_type` |
+
+The pipeline endpoint uses the `frame_template_*` prefix because it already has a `frame_template` field (preset key). The single-frame endpoints use the shorter `template_html` because they don't.
+
+## Tip — getting unstuck
+
+Every level has `--help`:
+
+```bash
+rf --help                     # top-level overview
+rf pipelines --help           # list of pipelines
+rf pipelines standard --help  # full option reference
+rf tts edge --help            # one specific command
+```
+
+## License
+
+Apache-2.0

package/dist/commands/create.js

@@ -66,6 +66,37 @@ async function resolveTextOrFile(input) {
     }
     return input;
 }
+/**
+ * Reference-image resolver. Accepts a public URL, a data: URI, or a local
+ * file path. Local files are base64-encoded into a data: URI so RelayX can
+ * receive them in a pure-JSON body (no upload endpoint needed on our side).
+ *
+ * Returns undefined when input is missing/blank so the caller can branch on
+ * "user actually provided this knob".
+ */
+async function resolveRefImage(input, flagName) {
+    if (input === undefined)
+        return undefined;
+    const t = input.trim();
+    if (!t)
+        return undefined;
+    if (/^https?:\/\//i.test(t) || t.startsWith("data:"))
+        return t;
+    const abs = path.resolve(t);
+    if (!fsSync.existsSync(abs)) {
+        throw new Error(`${flagName}: local file not found: ${abs}`);
+    }
+    const ext = path.extname(abs).toLowerCase();
+    const mime = ext === ".jpg" || ext === ".jpeg" ? "image/jpeg" :
+        ext === ".webp" ? "image/webp" :
+            ext === ".png" ? "image/png" :
+                null;
+    if (!mime) {
+        throw new Error(`${flagName}: unsupported extension ${ext} (use png/jpg/jpeg/webp)`);
+    }
+    const buf = await fs.readFile(abs);
+    return `data:${mime};base64,${buf.toString("base64")}`;
+}
 async function loadRecipe(recipePath) {
     const raw = await fs.readFile(recipePath, "utf-8");
     const parsed = JSON.parse(raw);
@@ -151,6 +182,8 @@ function optsToBody(opts) {
         out.image_model = opts.imageModel;
     if (opts.promptPrefix !== undefined)
         out.prompt_prefix = opts.promptPrefix;
+    if (opts.characterRef !== undefined)
+        out.character_ref = opts.characterRef;
     if (opts.voiceId !== undefined)
         out.voice_id = opts.voiceId;
     if (opts.ttsSpeed !== undefined)
@@ -304,9 +337,10 @@ export function registerCreate(program) {
         .option("--frame-template <keyOrPath>", "HTML frame template: preset key (e.g. 1080x1920/image_default.html) OR path to a local .html (auto-sent inline)")
         .option("--frame-template-size <wxh>", "size for inline HTML when the file lacks <meta template:width|height>, e.g. 1080x1920")
         .option("--frame-template-type <type>", "inline template type: image (default) | static | asset. Controls whether AI image generation runs per scene.")
-        .option("--image-model <id>", "RelayX image model (rx-image-z | rx-image-flux | rx-image-qwen)")
+        .option("--image-model <id>", "RelayX image model (rx-image-z | rx-image-flux | rx-image-qwen | rx-image-qwen-edit). Auto-switches to rx-image-qwen-edit when --character-ref is set.")
         .option("--prompt-prefix <text>", "raw style prefix prepended to every image prompt (overrides --style)")
         .option("--style <preset>", "image style preset — shortcut for --prompt-prefix; see 'Style presets' below")
+        .option("--character-ref <urlOrPath>", "reference image of the main character — locks identity across scenes. URL, data: URI, or local png/jpg/webp path (auto-encoded). Auto-enables rx-image-qwen-edit.")
         // --- Audio (TTS) ---
         .option("--voice-id <id>", "RelayX TTS voice id (default 专业解说); see `rf tts voices`")
         .option("--tts-speed <n>", "speech speed 0.5..2 (default 1.0)", parseFloat)
@@ -378,6 +412,10 @@ export function registerCreate(program) {
         "  # Pick a built-in style preset",
         '  rf create "..." --style cinematic',
         "",
+        "  # Cross-scene character consistency (auto-switches image model)",
+        '  rf create "主角小女孩的一天" --character-ref ./hero.png',
+        '  rf create "..." --character-ref https://example.com/hero.png',
+        "",
         "  # Recipe + replay last",
         "  rf create --recipe ./space.recipe.json",
         "  rf create --redo                       # replay last successful create",
@@ -438,6 +476,14 @@ export function registerCreate(program) {
         if (typeof body.script === "string") {
             body.script = await resolveTextOrFile(body.script);
         }
+        // Resolve character ref: local file path → data: URI (RelayX accepts
+        // both https:// and data: in image_urls). Done after layering so a
+        // recipe can carry the ref by path too.
+        const resolvedChar = await resolveRefImage(body.character_ref, "--character-ref");
+        if (resolvedChar !== undefined)
+            body.character_ref = resolvedChar;
+        else
+            delete body.character_ref;
         // Validate content mode
         const hasTopic = typeof body.topic === "string" && body.topic.trim().length > 0;
         const hasScript = typeof body.script === "string" && body.script.trim().length > 0;

package/dist/commands/images.js

@@ -1,6 +1,35 @@
+import fs from "node:fs/promises";
+import fsSync from "node:fs";
+import path from "node:path";
 import { post } from "../client.js";
 import { downloadTo } from "../utils/download.js";
 import { print, success } from "../utils/output.js";
+/**
+ * Reference image resolver. Accepts an https:// URL, a data: URI, or a local
+ * file path. Local files are read + base64-encoded into a data: URI so the
+ * server can forward them inside the JSON body — no file-hosting step needed.
+ */
+async function resolveImageRef(input) {
+    const t = input.trim();
+    if (!t)
+        throw new Error("--image: empty value");
+    if (/^https?:\/\//i.test(t) || t.startsWith("data:"))
+        return t;
+    const abs = path.resolve(t);
+    if (!fsSync.existsSync(abs)) {
+        throw new Error(`--image: local file not found: ${abs}`);
+    }
+    const ext = path.extname(abs).toLowerCase();
+    const mime = ext === ".jpg" || ext === ".jpeg" ? "image/jpeg" :
+        ext === ".webp" ? "image/webp" :
+            ext === ".png" ? "image/png" :
+                null;
+    if (!mime) {
+        throw new Error(`--image: unsupported extension ${ext} (use png/jpg/jpeg/webp)`);
+    }
+    const buf = await fs.readFile(abs);
+    return `data:${mime};base64,${buf.toString("base64")}`;
+}
 export function registerImages(program) {
     const images = program
         .command("images")
@@ -8,19 +37,30 @@ export function registerImages(program) {
         .helpOption("-h, --help", "show help");
     images
         .command("generate")
-        .description("Generate an image via RelayX")
+        .description("Generate an image via RelayX (text-to-image or ref-image, depending on model + --image)")
         .helpOption("-h, --help", "show help")
-        .requiredOption("-p, --prompt <text>", "text prompt")
-        .option("-m, --model <id>", "RelayX image model id (rx-image-z | rx-image-flux | rx-image-qwen)")
+        .requiredOption("-p, --prompt <text>", "text prompt. With --image, reference each ref via 「图 N」/「Image N」.")
+        .option("-m, --model <id>", "RelayX image model id (rx-image-z | rx-image-flux | rx-image-qwen | rx-image-qwen-edit)")
         .option("--width <n>", "image width", parseInt)
         .option("--height <n>", "image height", parseInt)
+        .option("-i, --image <urlOrPath>", "reference image, repeatable up to 3 times. URL / data: URI / local png/jpg/webp path (auto-base64'd). Index N-1 maps to 「图 N」in the prompt. Only honored by edit-capable SKUs (rx-image-qwen-edit).", (val, prev) => [...prev, val], [])
         .option("-o, --output <file>", "download first image to this local path")
         .option("--all-output <dir>", "download ALL generated images into this directory")
         .addHelpText("after", [
         "",
         "Examples:",
+        "  # plain text-to-image",
         "  reelforge images generate -p 'a cat' -m rx-image-flux --width 1024 --height 1024 -o cat.png",
-        "  reelforge images generate -p 'cyberpunk city, neon, 9:16' --width 1080 --height 1920 -o city.png",
+        "",
+        "  # ref-image edit (Qwen-Image-Edit), 1 ref local file",
+        "  reelforge images generate -m rx-image-qwen-edit \\",
+        "    -p '保持图1人物，把背景改成雪山黄昏' \\",
+        "    -i ./hero.png -o out.png",
+        "",
+        "  # ref-image edit, 2 refs (URL + local), 「图1」+「图2」",
+        "  reelforge images generate -m rx-image-qwen-edit \\",
+        "    -p '让图1的人物穿上图2的衣服' \\",
+        "    -i https://example.com/person.jpg -i ./outfit.png -o composite.png",
     ].join("\n"))
         .action(async (opts) => {
         const body = { prompt: opts.prompt };
@@ -30,6 +70,12 @@ export function registerImages(program) {
             body.width = opts.width;
         if (opts.height !== undefined)
             body.height = opts.height;
+        if (opts.image && Array.isArray(opts.image) && opts.image.length > 0) {
+            if (opts.image.length > 3) {
+                throw new Error(`--image accepts at most 3 refs (got ${opts.image.length})`);
+            }
+            body.image_urls = await Promise.all(opts.image.map(resolveImageRef));
+        }
         const r = await post("/api/v1/images/generate", body);
         if (opts.output && r.images?.[0]) {
             await downloadTo(r.images[0], opts.output);

package/package.json

@@ -1,51 +1,51 @@
-{
-  "name": "reelforge",
-  "version": "0.6.0",
-  "description": "CLI for ReelForge Studio — AI video engine. Installs as both `reelforge` and the short alias `rf`. Every REST API exposed as a command, with --help on every level.",
-  "license": "Apache-2.0",
-  "type": "module",
-  "bin": {
-    "reelforge": "./bin/reelforge.js",
-    "rf": "./bin/reelforge.js"
-  },
-  "files": [
-    "bin",
-    "dist",
-    "README.md"
-  ],
-  "engines": {
-    "node": ">=18.17"
-  },
-  "scripts": {
-    "build": "tsc -p tsconfig.json",
-    "dev": "tsc -p tsconfig.json --watch",
-    "typecheck": "tsc -p tsconfig.json --noEmit",
-    "clean": "rimraf dist",
-    "prepublishOnly": "npm run clean && npm run build"
-  },
-  "dependencies": {
-    "commander": "^12.1.0",
-    "kleur": "^4.1.5"
-  },
-  "devDependencies": {
-    "@types/node": "^20.14.0",
-    "rimraf": "^6.0.1",
-    "typescript": "^5.5.0"
-  },
-  "keywords": [
-    "reelforge",
-    "ai-video",
-    "relayx",
-    "tts",
-    "edge-tts",
-    "ffmpeg",
-    "playwright",
-    "cli"
-  ],
-  "repository": {
-    "type": "git",
-    "url": "https://github.com/puke3615/ReelForge.git",
-    "directory": "cli"
-  },
-  "homepage": "https://github.com/puke3615/ReelForge"
-}
+{
+  "name": "reelforge",
+  "version": "0.7.1",
+  "description": "CLI for ReelForge Studio — AI video engine. Installs as both `reelforge` and the short alias `rf`. Every REST API exposed as a command, with --help on every level.",
+  "license": "Apache-2.0",
+  "type": "module",
+  "bin": {
+    "reelforge": "./bin/reelforge.js",
+    "rf": "./bin/reelforge.js"
+  },
+  "files": [
+    "bin",
+    "dist",
+    "README.md"
+  ],
+  "engines": {
+    "node": ">=18.17"
+  },
+  "scripts": {
+    "build": "tsc -p tsconfig.json",
+    "dev": "tsc -p tsconfig.json --watch",
+    "typecheck": "tsc -p tsconfig.json --noEmit",
+    "clean": "rimraf dist",
+    "prepublishOnly": "npm run clean && npm run build"
+  },
+  "dependencies": {
+    "commander": "^12.1.0",
+    "kleur": "^4.1.5"
+  },
+  "devDependencies": {
+    "@types/node": "^20.14.0",
+    "rimraf": "^6.0.1",
+    "typescript": "^5.5.0"
+  },
+  "keywords": [
+    "reelforge",
+    "ai-video",
+    "relayx",
+    "tts",
+    "edge-tts",
+    "ffmpeg",
+    "playwright",
+    "cli"
+  ],
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/puke3615/ReelForge.git",
+    "directory": "cli"
+  },
+  "homepage": "https://github.com/puke3615/ReelForge"
+}