video-workbench/projects/2026-06-23-在路上/slides/s15/s15-v2-prompt.md

S-15 Slide Prompt V2

Status

• Status: ready for host-native generation.
• Slide: S-15.
• Version: s15-v2.
• Running mode: prompt and generate.
• Review mode: none.
• Overwrite allowed: no.

Based On Brief

• Brief: projects/2026-06-23-在路上/slides/s15/s15-v2-brief.md.
• Packet: projects/2026-06-23-在路上/tasks/2026-06-24-character-consistency/session-08/packets/s15-v2-slide-agent-packet.md.
• Slide function: road continues with no clear destination; continuation, not arrival.

Reference Inputs

• Style module: A, Landscape-Led Road.
• Image 1: projects/2026-06-23-在路上/intake/imgs/A.风景主导型蓝图.png.
  • Role: style/composition reference; landscape-led muddy road, cold open terrain, world before person.
  • SHA256: 3BC6333FAE372275F65E1DAB2B79E0EA2B4A67559A0E91A31A83079CF444BCC4.
• Image 2: projects/2026-06-23-在路上/visual-system/characters/character-slice-walk-back-v2-from-anchor.png.
  • Role: walking-back full-body character identity/body reference.
  • SHA256: 638DE8BC027C2756DAA819B6038B25BE3D9FEB5A26842ED651A5BEAF4415AB89.
• Image 3: projects/2026-06-23-在路上/visual-system/characters/character-anchor-v2-walk-back-reference-crop-tight.png.
  • Role: backpack, coat, boots, and upright walking-away posture reference.
  • SHA256: E375F2F29B627E94204F7ED42839DA8C8F67D8745AAFD41ECF0E17DCFB2B23BB.
• Generation path note: actual images are attached in visible context and must be used with the roles above. Local paths are traceability only.

Operation Notes

• Use a 16:9 horizontal cinematic documentary still.
• First read order: road/world before person.
• Traveler scale: about 1/12 to 1/8 of frame height.
• Landscape share: at least 80% of the frame.
• No s15-v2-review.md should be produced.

Advisor / Host-Native Record

• Skill: gpt-image-2.
• Fresh mode check command: node C:\Users\wangq\.agents\skills\gpt-image-2\scripts\check-mode.js --json.
• Fresh mode check result:

{
  "mode": "B-or-C",
  "recommendation": "host-or-advisor",
  "garden_mode_enabled": false,
  "has_api_key": false,
  "base_url": "https://api.openai.com/v1",
  "model": "gpt-image-2",
  "env_flag_value": "(unset)",
  "summary": "MODE B / C · 未启用 Garden。若宿主 Agent 自带图像工具（image_generation / dalle / mcp__*image* 等）→ MODE B：把 prompt 交给宿主出图。若宿主无图像工具 → MODE C：仅产出高质量 prompt 给用户。"
}

• Interpreted mode for this worker: Mode B, Host-Native, because the host provides image_gen.
• Garden/API expected: no.
• Skill/template/reference paths read:
  • C:\Users\wangq\.agents\skills\gpt-image-2\SKILL.md
  • C:\Users\wangq\.agents\skills\gpt-image-2\references\prompt-writing.md
  • C:\Users\wangq\.agents\skills\gpt-image-2\references\scenes-and-illustrations\concept-scene.md
  • C:\Users\wangq\.agents\skills\gpt-image-2\references\scenes-and-illustrations\minimalist-mood-scene.md
• Project/source files actually read:
  • AGENTS.md
  • docs/workflows/slide-task-orchestration.md
  • docs/workflows/slide-task-templates.md
  • projects/2026-06-23-在路上/project.md
  • projects/2026-06-23-在路上/execution-plan.md
  • projects/2026-06-23-在路上/slides/slides.md
  • projects/2026-06-23-在路上/visual-system/style/style-prompt-master.md
  • projects/2026-06-23-在路上/visual-system/characters/character-reference-registry.md
  • projects/2026-06-23-在路上/slides/s15/s15-v2-brief.md
  • projects/2026-06-23-在路上/tasks/2026-06-24-character-consistency/session-08/packets/s15-v2-slide-agent-packet.md

Final Prompt

Use the attached reference images with these roles:

• Image 1 is the landscape-led road style and composition reference.
• Image 2 is the walking-away full-body character identity/body reference.
• Image 3 is the tight walking-away backpack, coat, boots, and posture reference.

Create one photorealistic 16:9 horizontal cinematic documentary road-film still for S-15.

The image must read first as a long cold muddy road continuing through a large open wilderness with no clear destination. The road is the protagonist: wet ruts, dark puddles, rough earth, stones, sparse frozen grass, frost, low gray haze, and ambiguous open terrain. Let the road recede diagonally or horizontally into mist, out of frame, or into a low horizon without any arrival marker.

Place the traveler as a small secondary figure, about 1/12 to 1/8 of the frame height, moving away or three-quarter back. Preserve the identity cues from Image 2 and Image 3: long dark road-worn coat, dark backpack with worn straps and flap, boots, upright non-hunched continuation posture, slightly tired but capable walking direction. No readable face is needed; do not invent a new face.

Inherit style module A: landscape-led muddy road in a large cold wilderness; real documentary texture with restrained poetic realism; world larger than person, road longer than person. Use low saturation cold gray, wet blue-gray, earth brown, dark khaki, and coal-black mud. If any warmth appears, it must be weak and incidental, not a sunrise or emotional reward.

Composition requirements: landscape dominates at least 80% of the frame; road/world before person; no clean symmetrical poster road; no heroic centered backshot. The traveler should feel like a moving coordinate inside the terrain, not the subject of a travel portrait.

Texture requirements: damp mud, wet rut highlights, frost on grass and stones, gritty road surface, documentary grain, natural lens feel, restrained contrast, no glossy beauty.

Negative / Avoid

Avoid travel advertising, scenic tourism, motivational sunrise, broad sunset, beautiful postcard road, hero backshot, centered triumphant pose, clean commercial road, fashion photography, portrait MV, large character subject, readable face, fantasy portal, magic glow, vehicle, tent, camping objects, group reunion, text, logo, watermark, UI, cyberpunk, neon, sci-fi.

Output Intent

• Output file: projects/2026-06-23-在路上/slides/s15/s15-v2-image.png.
• Intended result: continuation image for Section C; road continues beyond certainty, no destination or resolution.