Skip to content

armgit / hypit

Public Imported from GitHub
0

Clone any viral video with AI agents. Not just a script, the whole workflow: swap the face, the words, the B-roll, ship 100 variants in one command, and get your 100M views.

Move Narratage skill and document Seedance kits

rponef395a2395313
diff --git a/.agent/skills/narratage/references/playbooks/craft/seedance-directing.md b/.agent/skills/narratage/references/playbooks/craft/seedance-directing.md
deleted file mode 100644
index fae62f1b..00000000
--- a/.agent/skills/narratage/references/playbooks/craft/seedance-directing.md
+++ /dev/null
@@ -1,14 +0,0 @@
-# Seedance directing
-
-The motion prompt is an increment over an authored reference frame.
-
-- Write one or two readable events, not frame-by-frame choreography: macro action plus micro eye,
- brow, smile, concern, or breath feedback.
-- Start from the reference pose, objects, emotion, camera, and space. Do not introduce a new person,
- prop, room, wardrobe, or angle.
-- Give physical motion envelopes: objects start in the hand or on the surface, travel a short distance,
- move slowly and naturally, never fly, bounce, teleport, detach, or multiply.
-- Tie speech actions to actual Script order. Quote only words present in the Script; otherwise use
- semantic phases such as opening, turn, and verdict.
-- Keep the face visible and hands away from it unless contact is required.
-- End every generated-video prompt with the English no-overlay sentence in `reference-vlm.md`.
diff --git a/.agent/skills/narratage/SKILL.md b/.agents/skills/narratage/SKILL.md
similarity index 91%
rename from .agent/skills/narratage/SKILL.md
rename to .agents/skills/narratage/SKILL.md
index b805a594..a330dc6a 100644
--- a/.agent/skills/narratage/SKILL.md
+++ b/.agents/skills/narratage/SKILL.md
@@ -13,6 +13,8 @@ Source, and `svml.runtime.json`/`.ts` Runtime Profile separate.
- Environment/setup or API keys → read `references/environment.md` and `references/credentials.md`.
- SVML authoring → read `references/quickstart.md`, then the linked authoritative page under
`docs/quickstart/`.
+- Seedance prompt assembly → read `references/seedance-kits.md`; reuse an official vendored Kit and
+ SVS Recipe before writing format scaffolding by hand.
- Production craft or format choice → read `references/playbooks/index.md` and
`references/playbooks/svml-mapping.md`, then only the relevant craft/format file.
- Reference-video reverse engineering → read `references/reference-vlm.md`; have VLM emit
diff --git a/.agent/skills/narratage/agents/openai.yaml b/.agents/skills/narratage/agents/openai.yaml
similarity index 100%
rename from .agent/skills/narratage/agents/openai.yaml
rename to .agents/skills/narratage/agents/openai.yaml
diff --git a/.agent/skills/narratage/references/credentials.md b/.agents/skills/narratage/references/credentials.md
similarity index 100%
rename from .agent/skills/narratage/references/credentials.md
rename to .agents/skills/narratage/references/credentials.md
diff --git a/.agent/skills/narratage/references/environment.md b/.agents/skills/narratage/references/environment.md
similarity index 89%
rename from .agent/skills/narratage/references/environment.md
rename to .agents/skills/narratage/references/environment.md
index 48d89708..987d6cbf 100644
--- a/.agent/skills/narratage/references/environment.md
+++ b/.agents/skills/narratage/references/environment.md
@@ -24,5 +24,5 @@ pnpm narratage services up svml.runtime.json
pnpm narratage doctor svml.runtime.json
```
-The repository owns skill discovery: `.agent/skills` is canonical, with repository-root
+The repository owns skill discovery: `.agents/skills` is canonical, with repository-root
`.codex/skills` and `.claude/skills` links. This is checkout layout, not a skill setup operation.
diff --git a/.agent/skills/narratage/references/playbooks/craft/b-roll.md b/.agents/skills/narratage/references/playbooks/craft/b-roll.md
similarity index 82%
rename from .agent/skills/narratage/references/playbooks/craft/b-roll.md
rename to .agents/skills/narratage/references/playbooks/craft/b-roll.md
index 55138dae..abf9da9e 100644
--- a/.agent/skills/narratage/references/playbooks/craft/b-roll.md
+++ b/.agents/skills/narratage/references/playbooks/craft/b-roll.md
@@ -1,5 +1,8 @@
# B-roll craft
+- Use the vendored `broll-v1` Kit with `text:Render`. Put `material-mode`, `story-shape`,
+ `edit-language`, `camera-language`, and `motion-intensity` in an SVS Recipe; connect only the
+ shot-specific micro-story through the required `story` slot.
- Choose the narrative scene before writing prompts. A useful product arc is **problem → failed
effort → product/mechanism → payoff**, but use the number of beats the story needs.
- Design each shot as a paired package: the image prompt establishes every visible fact; the video
diff --git a/.agent/skills/narratage/references/playbooks/craft/captions.md b/.agents/skills/narratage/references/playbooks/craft/captions.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/captions.md
rename to .agents/skills/narratage/references/playbooks/craft/captions.md
diff --git a/.agent/skills/narratage/references/playbooks/craft/image-prompt-style.md b/.agents/skills/narratage/references/playbooks/craft/image-prompt-style.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/image-prompt-style.md
rename to .agents/skills/narratage/references/playbooks/craft/image-prompt-style.md
diff --git a/.agent/skills/narratage/references/playbooks/craft/overlays.md b/.agents/skills/narratage/references/playbooks/craft/overlays.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/overlays.md
rename to .agents/skills/narratage/references/playbooks/craft/overlays.md
diff --git a/.agent/skills/narratage/references/playbooks/craft/persona-and-audio.md b/.agents/skills/narratage/references/playbooks/craft/persona-and-audio.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/persona-and-audio.md
rename to .agents/skills/narratage/references/playbooks/craft/persona-and-audio.md
diff --git a/.agent/skills/narratage/references/playbooks/craft/pip-overlay.md b/.agents/skills/narratage/references/playbooks/craft/pip-overlay.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/pip-overlay.md
rename to .agents/skills/narratage/references/playbooks/craft/pip-overlay.md
diff --git a/.agent/skills/narratage/references/playbooks/craft/production-gates.md b/.agents/skills/narratage/references/playbooks/craft/production-gates.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/production-gates.md
rename to .agents/skills/narratage/references/playbooks/craft/production-gates.md
diff --git a/.agent/skills/narratage/references/playbooks/craft/screen-demo.md b/.agents/skills/narratage/references/playbooks/craft/screen-demo.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/screen-demo.md
rename to .agents/skills/narratage/references/playbooks/craft/screen-demo.md
diff --git a/.agents/skills/narratage/references/playbooks/craft/seedance-directing.md b/.agents/skills/narratage/references/playbooks/craft/seedance-directing.md
new file mode 100644
index 00000000..566d6f02
--- /dev/null
+++ b/.agents/skills/narratage/references/playbooks/craft/seedance-directing.md
@@ -0,0 +1,25 @@
+# Seedance directing
+
+Separate the reusable format contract from per-take dialogue, action, story, or direction.
+
+- Select an official vendored Seedance Kit before composing prompt prose: `speaker-v1` for talking
+ heads, `broll-v1` for silent visual stories, `podcast-v1`, `call-v1`,
+ `street-interview-v1`, `motion-reference-v1`, or `camera-reference-v1`.
+- Render the Kit with `text:Render`; keep stable policy in an SVS Recipe and connect only dynamic
+ `dialogue`, `action`, `story`, or `direction` slots through `text:Set`. Do not duplicate the Kit's
+ reference, role, voice, camera, microphone, or hygiene blocks in a hand-written prompt.
+- Vendor the selected `.svs` into the video project's `./kits/` directory. Do not import it from a
+ separate Narratage checkout.
+
+- Write one or two readable events, not frame-by-frame choreography: macro action plus micro eye,
+ brow, smile, concern, or breath feedback.
+- Start from the reference pose, objects, emotion, camera, and space. Do not introduce a new person,
+ prop, room, wardrobe, or angle.
+- Give physical motion envelopes: objects start in the hand or on the surface, travel a short distance,
+ move slowly and naturally, never fly, bounce, teleport, detach, or multiply.
+- Tie speech actions to actual Script order. Quote only words present in the Script; otherwise use
+ semantic phases such as opening, turn, and verdict.
+- Keep the face visible and hands away from it unless contact is required.
+- For a prompt with no matching Kit, end it with the English no-overlay sentence in
+ `reference-vlm.md`. Official Kits already own their format-specific hygiene block; do not append a
+ duplicate.
diff --git a/.agent/skills/narratage/references/playbooks/craft/sfx.md b/.agents/skills/narratage/references/playbooks/craft/sfx.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/sfx.md
rename to .agents/skills/narratage/references/playbooks/craft/sfx.md
diff --git a/.agent/skills/narratage/references/playbooks/craft/visual-continuity.md b/.agents/skills/narratage/references/playbooks/craft/visual-continuity.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/craft/visual-continuity.md
rename to .agents/skills/narratage/references/playbooks/craft/visual-continuity.md
diff --git a/.agent/skills/narratage/references/playbooks/formats/asmr.md b/.agents/skills/narratage/references/playbooks/formats/asmr.md
similarity index 77%
rename from .agent/skills/narratage/references/playbooks/formats/asmr.md
rename to .agents/skills/narratage/references/playbooks/formats/asmr.md
index 0e4a4893..d2331027 100644
--- a/.agent/skills/narratage/references/playbooks/formats/asmr.md
+++ b/.agents/skills/narratage/references/playbooks/formats/asmr.md
@@ -2,6 +2,9 @@
Evidence level: thin.
+- Use the vendored `broll-v1` Kit. Select a practical material mode, a single readable story shape,
+ reference-locked camera language, and micro/readable motion in the Recipe; connect the exact
+ physical action through the `story` slot.
- Use close, simple, predictable physical actions: cutting, pouring, wiping, pressing, folding, or
rubbing. Keep motion micro/readable, not energetic or impact-heavy.
- Keep the shot continuous and camera close/locked. Rapid cuts or large camera moves break the
diff --git a/.agent/skills/narratage/references/playbooks/formats/mass-tarot.md b/.agents/skills/narratage/references/playbooks/formats/mass-tarot.md
similarity index 83%
rename from .agent/skills/narratage/references/playbooks/formats/mass-tarot.md
rename to .agents/skills/narratage/references/playbooks/formats/mass-tarot.md
index 4d011dd6..aac69e23 100644
--- a/.agent/skills/narratage/references/playbooks/formats/mass-tarot.md
+++ b/.agents/skills/narratage/references/playbooks/formats/mass-tarot.md
@@ -2,6 +2,8 @@
Evidence level: thin.
+- Use the vendored `broll-v1` Kit for card/table motion. Keep shared material, story shape, camera,
+ edit, and motion policy in a Recipe; connect each reveal action through the `story` slot.
- Preserve the four-part structure: question → three face-down cards → pick 1/2/3 prompt → individual
reveals and interpretations.
- Frame interpretation as reflection, not guaranteed fortune-telling. Offer themes/directions, not
diff --git a/.agent/skills/narratage/references/playbooks/formats/mixcut.md b/.agents/skills/narratage/references/playbooks/formats/mixcut.md
similarity index 83%
rename from .agent/skills/narratage/references/playbooks/formats/mixcut.md
rename to .agents/skills/narratage/references/playbooks/formats/mixcut.md
index 324bd329..8ae27fab 100644
--- a/.agent/skills/narratage/references/playbooks/formats/mixcut.md
+++ b/.agents/skills/narratage/references/playbooks/formats/mixcut.md
@@ -1,5 +1,7 @@
# Silent visual mixcut format
+- Render each generated beat through the vendored `broll-v1` Kit. Reuse an SVS Recipe for the shared
+ material/edit/camera language and vary only the beat-specific `story` Text when appropriate.
- Each beat is a reference frame plus a silent generated motion clip. Keep clips 4–15 seconds, use
distinct roles (hook, rhythm, detail/effect, conversion/payoff), and assemble ordered
`media-track:Item` inputs deliberately.
diff --git a/.agent/skills/narratage/references/playbooks/formats/ranking-listicle.md b/.agents/skills/narratage/references/playbooks/formats/ranking-listicle.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/formats/ranking-listicle.md
rename to .agents/skills/narratage/references/playbooks/formats/ranking-listicle.md
diff --git a/.agent/skills/narratage/references/playbooks/formats/scenario-call.md b/.agents/skills/narratage/references/playbooks/formats/scenario-call.md
similarity index 69%
rename from .agent/skills/narratage/references/playbooks/formats/scenario-call.md
rename to .agents/skills/narratage/references/playbooks/formats/scenario-call.md
index f72d27e5..4ca51caa 100644
--- a/.agent/skills/narratage/references/playbooks/formats/scenario-call.md
+++ b/.agents/skills/narratage/references/playbooks/formats/scenario-call.md
@@ -1,7 +1,9 @@
# Scenario-call format
-This is an editorial recipe, not a native call generator. Use two `seedance:ReferenceVideo` outputs
-or supplied video assets, then compose main/inset views with `media-track:Track` and explicit Frames.
+Use the vendored data-only `call-v1` Kit; it is a reusable Text Template, not a new execution node.
+Render it with `text:Render` plus an SVS Recipe, then feed one `seedance:ReferenceVideo` with the two
+final reversed call-layout images and two ordered voice references. Use supplied media and
+`media-track:Track` only when the call must be assembled from independent real feeds.
Evidence level: thin.
diff --git a/.agent/skills/narratage/references/playbooks/formats/street-interview.md b/.agents/skills/narratage/references/playbooks/formats/street-interview.md
similarity index 61%
rename from .agent/skills/narratage/references/playbooks/formats/street-interview.md
rename to .agents/skills/narratage/references/playbooks/formats/street-interview.md
index 7f1aeb1b..68df6bde 100644
--- a/.agent/skills/narratage/references/playbooks/formats/street-interview.md
+++ b/.agents/skills/narratage/references/playbooks/formats/street-interview.md
@@ -1,11 +1,15 @@
# Street-interview format
-This is an editorial recipe, not a dedicated Narratage generation component. Compose it with a
-`seedance:ReferenceVideo` per authored take (or supplied takes), explicit image/audio references,
-and a shared `speech:Spine`; use `media-track:Track` for PIP/reaction layout.
+Use the vendored data-only `street-interview-v1` Kit; it is a reusable Text Template, not a new
+execution node. Render it with `text:Render` plus an SVS Recipe, then feed the result to one
+`seedance:ReferenceVideo` with one complete-scene image and two ordered voice references. Use a
+shared `speech:Spine` for take assembly and `media-track:Track` only for separate editorial overlays.
- Establish one shared scene reference showing interviewer, guest, microphone, distance, and street
context. Establish two short clean voice references and a Script with explicit speaker cues.
+- Use `A:` for the interviewer/first audio reference and `B:` for the guest/second audio reference.
+ Connect dialogue and optional take-specific action through the Kit slots instead of rebuilding its
+ role, microphone, scene, and no-overlay prompt blocks.
- Keep one locked or softly handheld two-person view. Drift may be slight, but do not cut to a new
angle or reconstruct the room. The active speaker is identified by mouth and microphone; the listener
remains visibly alive and silent.
diff --git a/.agent/skills/narratage/references/playbooks/formats/talking-head.md b/.agents/skills/narratage/references/playbooks/formats/talking-head.md
similarity index 82%
rename from .agent/skills/narratage/references/playbooks/formats/talking-head.md
rename to .agents/skills/narratage/references/playbooks/formats/talking-head.md
index 264e3c19..e6419f2a 100644
--- a/.agent/skills/narratage/references/playbooks/formats/talking-head.md
+++ b/.agents/skills/narratage/references/playbooks/formats/talking-head.md
@@ -4,6 +4,8 @@
segments, each planned for 4–15 seconds with an 8–12 second target.
- Keep one visual identity across A-roll: same face reference, room, wardrobe, and lighting. Use
B-roll for changes, not to drift the presenter base.
+- Use the vendored `speaker-v1` Kit with `text:Render`; select composition, camera, edit,
+ performance, and gesture through an SVS Recipe, then connect dialogue/action as Text edges.
- Use `seedance:ReferenceVideo` for face/image/audio references; use `seedance:TextVideo` without an
identity reference. Keep dialogue and action as separate `text:Value` inputs.
- Build one `speech:Spine` and one `whisperx:Alignment`. Add `caption:Program` +
diff --git a/.agent/skills/narratage/references/playbooks/formats/two-person-podcast.md b/.agents/skills/narratage/references/playbooks/formats/two-person-podcast.md
similarity index 62%
rename from .agent/skills/narratage/references/playbooks/formats/two-person-podcast.md
rename to .agents/skills/narratage/references/playbooks/formats/two-person-podcast.md
index a06766b7..1cb6bd39 100644
--- a/.agent/skills/narratage/references/playbooks/formats/two-person-podcast.md
+++ b/.agents/skills/narratage/references/playbooks/formats/two-person-podcast.md
@@ -1,13 +1,17 @@
# Two-person podcast format
-This is an editorial recipe, not a native podcast surface. Compose one or more
-`seedance:ReferenceVideo` takes with two final camera-position references and explicit audio
-references; use ordered `speech:Take` inputs and peer Tracks for assembly.
+Use the vendored data-only `podcast-v1` Kit; it is a reusable Text Template, not a new execution
+node. Render it with `text:Render` plus an SVS Recipe, then feed the result to
+`seedance:ReferenceVideo` with two final camera-position images and two ordered voice references.
+Use ordered `speech:Take` inputs and peer Tracks for assembly.
- Prepare two final camera-position reference images, not isolated portraits. They establish room,
lighting, clothing, seating, and mutually consistent reverse views.
- Prefix every Script line with an explicit speaker cue. Only the active speaker talks; the other host
remains present with breathing, listening, and reaction.
+- Use `A:`/`B:` dialogue and connect optional take-specific action through the Kit. Keep framing,
+ edit language, pacing, performance, reaction, and gesture in the Recipe rather than rewriting the
+ prompt scaffold.
- Choose rhythm deliberately: follow the speaker for measured dialogue or cut to a meaningful listener
reaction. Framing changes are crops/pushes inside locked A/B views, never invented cameras/locations.
- Use one `speech:Spine`, at most one `caption:Program` + `caption-fine:Track`, and `text:Track`/
diff --git a/.agent/skills/narratage/references/playbooks/formats/voiceover-desk-demo.md b/.agents/skills/narratage/references/playbooks/formats/voiceover-desk-demo.md
similarity index 100%
rename from .agent/skills/narratage/references/playbooks/formats/voiceover-desk-demo.md
rename to .agents/skills/narratage/references/playbooks/formats/voiceover-desk-demo.md
diff --git a/.agent/skills/narratage/references/playbooks/index.md b/.agents/skills/narratage/references/playbooks/index.md
similarity index 94%
rename from .agent/skills/narratage/references/playbooks/index.md
rename to .agents/skills/narratage/references/playbooks/index.md
index bb7a7806..c345200d 100644
--- a/.agent/skills/narratage/references/playbooks/index.md
+++ b/.agents/skills/narratage/references/playbooks/index.md
@@ -8,7 +8,8 @@ private renderer internals or an analysis-data contract.
1. Identify the format: talking head, street interview, two-person podcast, voiceover desk demo, or
silent mixcut. Treat format files as editorial recipes; Narratage does not provide a native
generation surface for every format.
-2. Apply universal gates: `production-gates`, `visual-continuity`, and `seedance-directing`.
+2. Choose an official Seedance Kit when one matches, then apply universal gates:
+ `production-gates`, `visual-continuity`, and `seedance-directing`.
3. Add focused craft: `image-prompt-style`, `b-roll`, `overlays`, `captions`, `persona-and-audio`,
or `screen-demo`.
4. Read `svml-mapping.md`, then map the result to actual Narratage surfaces: Script, Seedance `TextVideo`/`FrameVideo`/
diff --git a/.agent/skills/narratage/references/playbooks/svml-mapping.md b/.agents/skills/narratage/references/playbooks/svml-mapping.md
similarity index 92%
rename from .agent/skills/narratage/references/playbooks/svml-mapping.md
rename to .agents/skills/narratage/references/playbooks/svml-mapping.md
index 3a84d518..101395d7 100644
--- a/.agent/skills/narratage/references/playbooks/svml-mapping.md
+++ b/.agents/skills/narratage/references/playbooks/svml-mapping.md
@@ -7,6 +7,7 @@ not aliases for the old canvas/JSON workflow.
|---|---|
| Story and dialogue | `Script`, `Segment`, `Role Cue`, `Selection`, `Moment` |
| Prompt or reusable copy | `text:Value` and `text:Render` (backed by Text Templates) |
+| Reusable Seedance prompt contract | vendored `speaker-v1`, `broll-v1`, `podcast-v1`, `call-v1`, `street-interview-v1`, `motion-reference-v1`, or `camera-reference-v1` Kit + SVS Recipe |
| Generated image | `gpt:Image`, `nano:Image`/`nano:ProImage`, or `seedream:TextImage`/`ReferenceImage` |
| Text-to-video | `seedance:TextVideo` |
| First-frame-to-video | `seedance:FrameVideo` |
diff --git a/.agent/skills/narratage/references/quickstart.md b/.agents/skills/narratage/references/quickstart.md
similarity index 87%
rename from .agent/skills/narratage/references/quickstart.md
rename to .agents/skills/narratage/references/quickstart.md
index 5e730685..a2582ff7 100644
--- a/.agent/skills/narratage/references/quickstart.md
+++ b/.agents/skills/narratage/references/quickstart.md
@@ -7,6 +7,7 @@ The repository docs are authoritative:
| Script semantics | `docs/quickstart/script.md` |
| SVS Recipes | `docs/quickstart/styles.md` |
| Media and Seedance | `docs/quickstart/generation.md` |
+| Reusable Seedance Prompt Kits | `packages/seedance-kits/README.md` and the selected file under `packages/seedance-kits/kits/` |
| Speech Spine and SemanticMap | `docs/quickstart/timing.md` |
| Caption, Media, Text, Audio Tracks | `docs/quickstart/tracks.md` and the `@narratage/audio-track` package README |
| Film and rendering | `docs/quickstart/composition.md` |
diff --git a/.agent/skills/narratage/references/reference-vlm.md b/.agents/skills/narratage/references/reference-vlm.md
similarity index 86%
rename from .agent/skills/narratage/references/reference-vlm.md
rename to .agents/skills/narratage/references/reference-vlm.md
index 149d447e..9b12cc1d 100644
--- a/.agent/skills/narratage/references/reference-vlm.md
+++ b/.agents/skills/narratage/references/reference-vlm.md
@@ -11,6 +11,9 @@ overlay JSON or introduce a JSON-to-SVML conversion step.
provided media; generated live-action PIP/B-roll may be real generated media.
- Anchor recurring people/products/locations at the clearest shot and reuse references topologically.
- Listen to audio for voice design. Transcribe speech, not burned-in captions.
+- Before hand-writing a Seedance prompt, select an official vendored Kit when the format matches:
+ `speaker-v1`, `broll-v1`, `podcast-v1`, `call-v1`, `street-interview-v1`,
+ `motion-reference-v1`, or `camera-reference-v1`. Supply only Recipe axes and dynamic slots.
- Map transcript to Script; generated media to `seedance:TextVideo`/`FrameVideo`/`ReferenceVideo`;
supplied assets to `media:Image`/`media:Audio` or `media-track:Item`; timing to `speech:Spine` and
`whisperx:Alignment`; overlays to `caption:Program` + `caption-fine:Track`, `media-track:Track`,
diff --git a/.agents/skills/narratage/references/seedance-kits.md b/.agents/skills/narratage/references/seedance-kits.md
new file mode 100644
index 00000000..8d767dac
--- /dev/null
+++ b/.agents/skills/narratage/references/seedance-kits.md
@@ -0,0 +1,26 @@
+# Seedance Prompt Kits
+
+Narratage ships seven data-only English Text Templates under `packages/seedance-kits/kits/`. They
+assemble recurring prompt contracts without adding execution nodes or hiding media/model choices.
+
+| Format | Kit | Required dynamic slots | Ordered references |
+|---|---|---|---|
+| Talking head | `speaker-v1.svs` | `dialogue`; optional `action` | image 1 = speaker/scene, audio 1 = voice |
+| Silent B-roll | `broll-v1.svs` | `story` | one or more authored images |
+| Two-person podcast | `podcast-v1.svs` | `dialogue`; optional `action` | images 1/2 = A/B views, audio 1/2 = A/B voices |
+| Video call | `call-v1.svs` | `dialogue`; optional `action` | images 1/2 = reversed call layouts, audio 1/2 = A/B voices |
+| Street interview | `street-interview-v1.svs` | `dialogue`; optional `action` | image 1 = complete scene, audio 1/2 = interviewer/guest |
+| Motion transfer | `motion-reference-v1.svs` | optional `direction` | image 1 = subject, video 1 = motion reference |
+| Camera transfer | `camera-reference-v1.svs` | optional `direction` | image 1 = subject, video 1 = camera reference |
+
+## Required workflow
+
+1. Read `packages/seedance-kits/README.md` and the selected Kit file for exact axes and defaults.
+2. Copy only that `.svs` file into the video project's `./kits/` directory; import the vendored copy.
+3. Put stable axis choices in `studio.svs` as a named Recipe.
+4. Use `text:Render` with the Kit Template and Recipe. Connect dynamic slots with `text:Set`.
+5. Feed the rendered Text to the appropriate low-level Seedance Surface. Keep duration, resolution,
+ model, and every image/audio/video reference explicit on the graph.
+
+Do not hand-write or duplicate a Kit's fixed reference, role, voice, microphone, camera, or hygiene
+blocks. Write a freeform English prompt only when none of the seven Kit contracts matches.
diff --git a/.agent/skills/narratage/scripts/check-credentials.mjs b/.agents/skills/narratage/scripts/check-credentials.mjs
similarity index 100%
rename from .agent/skills/narratage/scripts/check-credentials.mjs
rename to .agents/skills/narratage/scripts/check-credentials.mjs
diff --git a/.agent/skills/narratage/scripts/check-environment.mjs b/.agents/skills/narratage/scripts/check-environment.mjs
similarity index 100%
rename from .agent/skills/narratage/scripts/check-environment.mjs
rename to .agents/skills/narratage/scripts/check-environment.mjs
diff --git a/.gitignore b/.gitignore
index a8960f8a..8608dd74 100644
--- a/.gitignore
+++ b/.gitignore
@@ -10,7 +10,9 @@ coverage/
# Local editor and agent state is never part of the distribution.
.claude/
.codex/
-.agents/
+.agents/*
+!.agents/skills/
+!.agents/skills/**
.idea/
.vscode/
diff --git a/docs/quickstart/generation.md b/docs/quickstart/generation.md
index 3c7312f5..b63e68ae 100644
--- a/docs/quickstart/generation.md
+++ b/docs/quickstart/generation.md
@@ -16,7 +16,7 @@ Every component shown here must be imported by its package specifier before use:
<import as="estimate" from="@narratage/estimate@1"/>
<import as="text" from="@narratage/text@1"/>
<import as="seedance" from="@narratage/seedance@1"/>
-<import as="speaker-kit" source="../../packages/seedance-kits/kits/speaker-v1.svs"/>
+<import as="speaker-kit" source="./kits/speaker-v1.svs"/>
```
## media:Image
@@ -189,10 +189,14 @@ Runtime Endpoints for these exact Needs; neither changes the author graph.
generic `text:Render` to produce the prompt, then connect that Text and the real media references to
the low-level Seedance Surface.
+Copy only the selected Kit `.svs` files into the video project's `./kits/` directory. Import the
+vendored project copy so the Kit bytes remain inside the Source Closure; do not reach back into a
+Narratage repository checkout from project source.
+
```svml
<import as="text" from="@narratage/text@1"/>
<import as="seedance" from="@narratage/seedance@1"/>
-<import as="broll-kit" source="../../packages/seedance-kits/kits/broll-v1.svs"/>
+<import as="broll-kit" source="./kits/broll-v1.svs"/>
<text:Render id="demo-prompt"
template={broll-kit.broll-v1}
@@ -221,6 +225,41 @@ references plus two ordered audio references; Street Interview expects one scene
ordered voices; the two reference-transfer templates expect one subject image and one reference
video. Those shapes are visible in `seedance:ReferenceVideo`, not hidden in Kit execution code.
+## Street-interview prompt assembly
+
+Use `street-interview-v1` for the stable scene, role, microphone, voice and no-overlay contracts.
+Select framing, edit, pacing, performance, reaction and gesture through an SVS Recipe; provide only
+the authored dialogue and optional per-take action as dynamic Text edges:
+
+```svml
+<import as="text" from="@narratage/text@1"/>
+<import as="seedance" from="@narratage/seedance@1"/>
+<import as="interview-kit" source="./kits/street-interview-v1.svs"/>
+
+<text:Value id="interview-action">
+ Let the guest pause briefly before the answer; keep the microphone handoff natural.
+</text:Value>
+
+<text:Render id="interview-prompt"
+ template={interview-kit.street-interview-v1}
+ recipe={studio.interview.street}>
+ <text:Set name="dialogue" text={story.segment.interview.dialogue}/>
+ <text:Set name="action" text={interview-action}/>
+</text:Render>
+
+<seedance:ReferenceVideo id="interview-take" model="mini"
+ prompt={interview-prompt} duration={interview-duration.duration}
+ resolution="720p" aspect-ratio="9:16" generate-audio="true">
+ <seedance:Reference image={interview-scene}/>
+ <seedance:Reference audio={interviewer-voice}/>
+ <seedance:Reference audio={guest-voice}/>
+</seedance:ReferenceVideo>
+```
+
+The dialogue uses explicit `A:`/`B:` order: A is the interviewer and maps to the first audio
+reference; B is the guest and maps to the second. The Kit owns the reusable English scaffold, so do
+not duplicate it in a hand-written prompt.
+
## Talking-head prompt assembly
Talking-head authoring does not need a special executable component. The data-only `speaker-v1`
@@ -230,7 +269,7 @@ module. The result enters Seedance through the same explicit `prompt` edge as an
```svml
<import as="text" from="@narratage/text@1"/>
<import as="seedance" from="@narratage/seedance@1"/>
-<import as="speaker-kit" source="../../packages/seedance-kits/kits/speaker-v1.svs"/>
+<import as="speaker-kit" source="./kits/speaker-v1.svs"/>
<text:Value id="hook-action">
Begin with urgent direct eye contact, then let the final admission land more quietly.
@@ -266,7 +305,7 @@ A two-take setup with estimated durations feeding explicit Text assembly and See
<import as="text" from="@narratage/text@1"/>
<import as="seedance" from="@narratage/seedance@1"/>
<import as="studio" source="./studio.svs"/>
-<import as="speaker-kit" source="../../packages/seedance-kits/kits/speaker-v1.svs"/>
+<import as="speaker-kit" source="./kits/speaker-v1.svs"/>
<media:Image id="presenter-clean" src="./assets/presenter-clean.png"/>
<media:Image id="presenter-alt" src="./assets/presenter-alt.png"/>
diff --git a/docs/quickstart/styles.md b/docs/quickstart/styles.md
index f3754405..f6978bd6 100644
--- a/docs/quickstart/styles.md
+++ b/docs/quickstart/styles.md
@@ -318,7 +318,6 @@ transfer Kits remain separate from Seedance execution.
```svs
broll.product-demo {
material-mode: product-beauty;
- reference-plan: scene-plus-refs;
story-shape: process-demo;
edit-language: insert-cutaway;
camera-language: product-macro;
@@ -326,6 +325,24 @@ broll.product-demo {
}
```
+Street Interview, Podcast and Call use the same Recipe mechanism rather than hand-written fixed
+prompt prose. For example:
+
+```svs
+interview.street {
+ framing: soft-handheld;
+ edit-language: pause-trim;
+ pacing: compact;
+ performance: natural-street;
+ reaction: active;
+ gesture: natural;
+}
+```
+
+`street-interview-v1` reads those six axes. `podcast-v1` and `call-v1` read the corresponding
+`framing`, `edit-language`, `pacing`, `performance`, `reaction` and `gesture` axes with their own
+finite values. The selected Kit file is the authority for allowed values and defaults.
+
Model, resolution, duration and reference media are not template policy. They stay on the exact
model Surface and graph edges.
diff --git a/docs/quickstart/tracks.md b/docs/quickstart/tracks.md
index cc6ad348..afb753e8 100644
--- a/docs/quickstart/tracks.md
+++ b/docs/quickstart/tracks.md
@@ -96,8 +96,8 @@ while a Dual Text Atom remains one indivisible visible unit. Fine never guesses
`active-box` is independently `off`, `current` or `trail`. `active-box-continuity: isolated` paints
one capsule per activated Atom; `joined` turns a trail into one ordered prefix whose background is
-continuous on each real browser line. Thus trail-colored text with a current-only Pill, the original
-Twinit behavior, is one Recipe—not a second renderer.
+continuous on each real browser line. Thus trail-colored text with a current-only Pill is one
+Recipe—not a second renderer.
Fine wraps only between complete Atoms and never clips author text. It intentionally has no
`max-lines`; use Cue bounds, Track width and font size to control density.
diff --git a/docs/zh/quickstart/generation.md b/docs/zh/quickstart/generation.md
index 224567aa..5df76e51 100644
--- a/docs/zh/quickstart/generation.md
+++ b/docs/zh/quickstart/generation.md
@@ -15,7 +15,7 @@ description: 声明媒体资源并使用 Seedance 生成视频。
<import as="estimate" from="@narratage/estimate@1"/>
<import as="text" from="@narratage/text@1"/>
<import as="seedance" from="@narratage/seedance@1"/>
-<import as="speaker-kit" source="../../packages/seedance-kits/kits/speaker-v1.svs"/>
+<import as="speaker-kit" source="./kits/speaker-v1.svs"/>
```
## media:Image
@@ -177,10 +177,13 @@ Endpoint,不会改变作者图。
`@narratage/seedance-kits` 包含七个纯数据 Text Template。Kit 不是模型包装器:先用通用
`text:Render` 生成 prompt,再把该 Text 与真实媒体引用显式接入低层 Seedance Surface。
+只把项目实际选择的 Kit `.svs` 文件复制进视频项目的 `./kits/` 目录,并导入这份项目内
+副本,使 Kit 字节保持在 Source Closure 内;项目源码不要反向引用 Narratage 仓库 checkout。
+
```svml
<import as="text" from="@narratage/text@1"/>
<import as="seedance" from="@narratage/seedance@1"/>
-<import as="broll-kit" source="../../packages/seedance-kits/kits/broll-v1.svs"/>
+<import as="broll-kit" source="./kits/broll-v1.svs"/>
<text:Render id="demo-prompt"
template={broll-kit.broll-v1}
@@ -201,6 +204,39 @@ action 和 extra 仍经由 `text:Set` 作为图边进入。现有模板包括 `s
`call-v1`、`street-interview-v1`、`motion-reference-v1`、`camera-reference-v1`。
媒体数量与顺序仍清楚地写在 `seedance:ReferenceVideo` 里,不藏进 Kit 代码。
+## 街访 Prompt 组装
+
+使用 `street-interview-v1` 复用场景、角色、麦克风、音色和无叠加文字契约。构图、剪辑、
+节奏、表演、反应与手势由 SVS Recipe 选择;动态输入只保留对白与可选的单段动作:
+
+```svml
+<import as="text" from="@narratage/text@1"/>
+<import as="seedance" from="@narratage/seedance@1"/>
+<import as="interview-kit" source="./kits/street-interview-v1.svs"/>
+
+<text:Value id="interview-action">
+ Let the guest pause briefly before the answer; keep the microphone handoff natural.
+</text:Value>
+
+<text:Render id="interview-prompt"
+ template={interview-kit.street-interview-v1}
+ recipe={studio.interview.street}>
+ <text:Set name="dialogue" text={story.segment.interview.dialogue}/>
+ <text:Set name="action" text={interview-action}/>
+</text:Render>
+
+<seedance:ReferenceVideo id="interview-take" model="mini"
+ prompt={interview-prompt} duration={interview-duration.duration}
+ resolution="720p" aspect-ratio="9:16" generate-audio="true">
+ <seedance:Reference image={interview-scene}/>
+ <seedance:Reference audio={interviewer-voice}/>
+ <seedance:Reference audio={guest-voice}/>
+</seedance:ReferenceVideo>
+```
+
+对白使用明确的 `A:`/`B:` 顺序:A 是采访者并绑定第一段音频参考,B 是受访者并绑定第二段。
+固定英文 Prompt 骨架由 Kit 负责,不要在手写 Prompt 中重复一遍。
+
## 口播 Prompt 组装
口播创作不需要一个特殊的可执行组件。数据化的 `speaker-v1` Template、项目 Recipe 与每段的
@@ -209,7 +245,7 @@ dialogue/action 由普通 Text 模块组装,结果再像其他生成任务一
```svml
<import as="text" from="@narratage/text@1"/>
<import as="seedance" from="@narratage/seedance@1"/>
-<import as="speaker-kit" source="../../packages/seedance-kits/kits/speaker-v1.svs"/>
+<import as="speaker-kit" source="./kits/speaker-v1.svs"/>
<text:Value id="hook-action">
先紧迫地直视镜头,最后一句收住声音。
@@ -243,7 +279,7 @@ dialogue/action 由普通 Text 模块组装,结果再像其他生成任务一
<import as="text" from="@narratage/text@1"/>
<import as="seedance" from="@narratage/seedance@1"/>
<import as="studio" source="./studio.svs"/>
-<import as="speaker-kit" source="../../packages/seedance-kits/kits/speaker-v1.svs"/>
+<import as="speaker-kit" source="./kits/speaker-v1.svs"/>
<media:Image id="presenter-clean" src="./assets/presenter-clean.png"/>
<media:Image id="presenter-alt" src="./assets/presenter-alt.png"/>
diff --git a/docs/zh/quickstart/styles.md b/docs/zh/quickstart/styles.md
index a68a595c..094594d0 100644
--- a/docs/zh/quickstart/styles.md
+++ b/docs/zh/quickstart/styles.md
@@ -310,7 +310,6 @@ Street Interview 与参考迁移 Kit 可以保持为纯数据,而不进入 See
```svs
broll.product-demo {
material-mode: product-beauty;
- reference-plan: scene-plus-refs;
story-shape: process-demo;
edit-language: insert-cutaway;
camera-language: product-macro;
@@ -318,6 +317,23 @@ broll.product-demo {
}
```
+Street Interview、Podcast 与 Call 使用同一套 Recipe 机制,不需要手写固定 Prompt。例如:
+
+```svs
+interview.street {
+ framing: soft-handheld;
+ edit-language: pause-trim;
+ pacing: compact;
+ performance: natural-street;
+ reaction: active;
+ gesture: natural;
+}
+```
+
+`street-interview-v1` 读取这六个轴。`podcast-v1` 与 `call-v1` 读取同名的 `framing`、
+`edit-language`、`pacing`、`performance`、`reaction` 和 `gesture` 轴,但使用各自的有限值。
+允许值和默认值以所选 Kit 文件为准。
+
模型、分辨率、时长和参考媒体不是模板策略;它们继续存在于精确模型 Surface 与显式图边中。
## 精确字体声明
diff --git a/packages/seedance-kits/README.md b/packages/seedance-kits/README.md
index f021457c..65b31455 100644
--- a/packages/seedance-kits/README.md
+++ b/packages/seedance-kits/README.md
@@ -58,6 +58,6 @@ The seven templates are:
- `motion-reference-v1`: preserve the subject and transfer body motion only;
- `camera-reference-v1`: preserve the subject and transfer camera language only.
-Speaker, B-roll, Podcast, Call and Street Interview retain useful orthogonal prompt axes from the legacy implementation. Historical aliases that emitted
-identical prose were removed. Speed, trim, last-frame extraction and audio extraction are ordinary
-media operations and intentionally do not live in these templates.
+Speaker, B-roll, Podcast, Call and Street Interview expose finite orthogonal prompt axes through
+their Text Templates. Speed, trim, last-frame extraction and audio extraction are ordinary media
+operations and intentionally do not live in these templates.