T8mars/minimax-h3-prompt-skill-T8 / skills/recombine-conversation-subgroups
recombine-conversation-subgroups
VerifiedCreate original short videos using the 分组对话覆盖|两边接力后全桌合流 Creative DNA for MiniMax H3 or Seedance 2.0. Use when the user wants this causal, camera, motion, rhythm, or payoff structure with new subjects and surfaces.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 2.6 KB |
| agents/openai.yaml | config | 549 B |
| references/summary.md | markdown | 1.4 KB |
| references/template.md | markdown | 4.3 KB |
references/template.md
View on GitHub# 分组对话覆盖|两边接力后全桌合流 Template
## Mechanism
Establish one shared social geography by showing the first seated pair, hand visual speaking-turn ownership within that pair, cut across the table to a second pair for a distinct response phase, then return to a wider composition that recombines all participants and the same table anchors. Source dialogue meaning is unknown; only visible turn ownership and coverage order are reusable.
## Invariants
### inv-01
- Rule: Establish the first subgroup and shared table anchors before any cross-table cut.
- Purpose: Creates a stable social half-space.
- Failure if removed: Opening on singles hides group geography.
### inv-02
- Rule: Make speaking-turn ownership visibly readable within the first subgroup without assigning source meaning.
- Purpose: Defines the coverage relay.
- Failure if removed: Indistinct mouth and eye ownership makes the cut arbitrary.
### inv-03
- Rule: Cut to a second subgroup that faces back into the same shared space.
- Purpose: Expands the conversation while preserving geography.
- Failure if removed: A new unrelated location breaks the group.
### inv-04
- Rule: Give the second subgroup one distinct response phase and one listening witness.
- Purpose: Prevents equal-motion visual noise.
- Failure if removed: Both people performing at once hides ownership.
### inv-05
- Rule: Recombine all participants with the same spatial anchors in a held wide view.
- Purpose: Closes the coverage map.
- Failure if removed: Ending on a pair leaves the larger group unresolved.
## Variable Slots
| Slot | Constraint |
|---|---|
| lead_subject | An original stable subject or pair whose role is clear without source identity. |
| setting | A materially new setting with fixed spatial anchors and no source-specific location. |
| causal_carrier | One object, route, mark or device that visibly carries the causal chain. |
| state_change | Five ordered states with observable handoffs, not a decorative montage. |
| camera_grammar | A new proof-oriented information order that does not copy source shot boundaries. |
| payoff | A held final consequence that resolves or recontextualizes the setup. |
| sound_family | An original hierarchy of story-world cues supporting each visible change. |
## Required Human-Facing Anchors
- Establish the first subgroup and shared table anchors before any cross-table cut.
- Make speaking-turn ownership visibly readable within the first subgroup without assigning source meaning.
- Give the second subgroup one distinct response phase and one listening witness.
- Recombine all participants with the same spatial anchors in a held wide view.
## Anti-Copy Boundary
- source beach bar and four actors
- source clothing and cocktails
- source dialogue content
- the exact male-pair to female-pair coverage
## Transfer Tests
1. Four polar greenhouse engineers sit around one circular sample table: the first pair compare a cracked seed shell, the opposite pair inspect a moisture graph, one technician rotates a sample cup, and a final wide shows all four agreeing on one marked tray without dialogue.
- Preserved: inv-01, inv-02, inv-03, inv-04, inv-05
- Changed: lead_subject, setting, causal_carrier, state_change, camera_grammar, payoff, sound_family
2. Four stop-motion harbor puppets share a chart desk: two place route tiles, the opposite pair respond with tide markers, one lifts a buoy token, and the final wide recombines every puppet and the completed route.
- Preserved: inv-01, inv-02, inv-03, inv-04, inv-05
- Changed: lead_subject, setting, causal_carrier, state_change, camera_grammar, payoff, sound_family
## Model Skeletons
### MiniMax H3
integrated_multimodal_description: [new subject, environment, ordered actions, camera, light and final state preserving every anchor]
overall_soundscape: [diegetic ambience, action cues and causal payoff]
non_diegetic_music: [original music arc or N/A]
### Seedance 2.0
任务:[原创实例意图]
总时长约[4–15]秒。
主体定义:[稳定人物、物体与空间关系]
事件或镜头:[按因果顺序实现全部锚点]
整体风格与画质:[原创媒介、光影、色彩与声音]
约束:[连续性、物理逻辑、反复制与结尾状态]
Adapt shot count to the user's concept. Preserve the causal mechanism, not a fixed storyboard.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |