# S-23 Slide Prompt V3

## Status

- Status: ready for host-native generation.
- Slide/version: `S-23 / s23-v3`.
- Running mode: prompt and generate.
- Review: disabled; do not write `s23-v3-review.md`.
- Overwrite allowed: no; target output paths were checked absent before this file was written.

## Based On Brief

- Brief: `projects/2026-06-23-在路上/slides/s23/s23-v3-brief.md`
- Packet: `projects/2026-06-23-在路上/tasks/2026-06-24-character-consistency/session-11/packets/s23-v3-slide-agent-packet.md`
- Slide function: open the chorus by sinking the image, not igniting it; large muddy ground before the traveler.

## Reference Inputs

- Image 1: common style calibration board, `projects/2026-06-23-在路上/visual-system/style/style-anchor-board-v1.png`; role: style calibration, palette, avoid-drift.
- Image 2: A landscape-led blueprint, `projects/2026-06-23-在路上/intake/imgs/A.风景主导型蓝图.png`; role: landscape-led muddy road / wilderness composition reference.
- Image 3: distant walking-back silhouette reference, `projects/2026-06-23-在路上/visual-system/characters/character-slice-walk-back-v2-from-anchor.png`; role: distant back-view walking silhouette, backpack and long-coat continuity.

## Advisor / Host-Native Record

- Skill: `gpt-image-2`.
- Mode check command: `node C:\Users\wangq\.agents\skills\gpt-image-2\scripts\check-mode.js --json`.
- Mode result: `B-or-C`; effective mode for this worker: Mode B / host-native because the host image generation tool is available.
- Garden/API expected: no.
- Garden scripts called: no.
- External services called: no MiniMax MCP, no Minimax CLI, no external vision/review/image understanding, no web search, no CCPE, no GPT V2.
- Skill references used:
  - `C:\Users\wangq\.agents\skills\gpt-image-2\SKILL.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`

## Source Files Read

- `AGENTS.md`
- `docs/workflows/slide-task-orchestration.md`
- `docs/workflows/slide-task-templates.md`
- `projects/2026-06-23-在路上/tasks/2026-06-24-character-consistency/session-11/packets/s23-v3-slide-agent-packet.md`
- `projects/2026-06-23-在路上/slides/s23/s23-v3-brief.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/intake/阶段五-镜头清单骨架.md`

## Final Prompt

Create one horizontal 16:9 photorealistic cinematic documentary road-film still for slide S-23, using the attached visible references as follows: Image 1 for overall style calibration and restraint, Image 2 for the landscape-led muddy wilderness road composition, and Image 3 only for the distant walking-away silhouette, backpack, long dark coat, and steady walking posture.

Scene: a broad cold muddy field or muddy wilderness path after rain, with wet ruts, uneven mud, dark puddles, stones, frost traces, rough earth, sparse dead grass, and low gray haze. The muddy ground must dominate the first read before the traveler. The world is larger than the person, the road is longer than the person, and the traveler is only a small coordinate still moving.

Subject: one traveler viewed from behind, walking slowly but continuously through the mud. Keep him distant and unheroic, about 1/14 to 1/9 of the frame. Preserve the Image 3 read of a worn backpack, long dark road coat, dark trousers, boots, and walking-away silhouette, but do not make this a portrait or identity-detail shot.

Composition: wide landscape frame, environment first, muddy ground second, traveler third. Use a low-to-medium documentary road-film perspective, with the path or muddy ruts leading into the distance. The frame should feel heavy and quiet, opening the chorus by sinking into the mud rather than rising into triumph.

Light and color: cold gray, earth brown, dark khaki, coal black, wet dark blue-gray; low saturation, low contrast, natural overcast or weak pre-dawn light. No bright inspirational sky, no postcard sunrise, no warm heroic rim light. Any warmth should be extremely restrained or absent.

Texture: real wet mud, old road ruts, shallow puddles, grit, stones, dirty water, road-worn clothing, documentary realism. Poetry should come from the reality of the mud and the continued movement, not from fantasy effects.

Avoid: heroic pose, centered triumphant backshot, high-energy chorus look, travel advertising, clean scenic landscape, motivational poster, portrait emphasis, fashion photography, shoe advertisement, clean commercial road, broad sunrise or sunset, high saturation, magic glow, readable text, labels, logos, watermarks, UI.

Output intent: a single 16:9 still image, no text, no border, no collage, no multiple panels.