wantsong
/
video-workbench
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
video-workbench/projects/2026-06-23-在路上/slides/s07/s07-v2-prompt.md

6.5 KiB
Raw Blame History

S-07 Slide Prompt V2

Status

  • Status: ready for host-native generation.
  • Worker: Agent S-07.
  • Slide/version: S-07 / s07-v2.
  • Running mode: prompt and generate.
  • Review: none; do not write s07-v2-review.md.
  • Overwrite allowed: no.

Based On Brief

  • projects/2026-06-23-在路上/slides/s07/s07-v2-brief.md
  • projects/2026-06-23-在路上/tasks/2026-06-24-character-consistency/session-06/packets/s07-v2-slide-agent-packet.md

Reference Inputs

  • Style module: A｜清晨泥泞旷野：世界先于人出现.
  • Active style reference: projects/2026-06-23-在路上/intake/imgs/A.风景主导型蓝图.png as landscape-led road composition reference.
  • Style calibration reference: projects/2026-06-23-在路上/visual-system/style/style-anchor-board-v1.png as continuity reference only.
  • Image 1: projects/2026-06-23-在路上/visual-system/characters/character-slice-walk-back-v2-from-anchor.png as identity/body continuity reference for walking away, long dark coat, old backpack, worn boots, and compact road-worn silhouette.
  • Image 2: projects/2026-06-23-在路上/visual-system/characters/character-slice-walk-side-v2-from-anchor.png as supporting walking side-posture reference for stride, coat fall, backpack placement, and boot profile.
  • Generation path note: Image 1 and Image 2 must be loaded into visible context before host image generation. Local paths are traceability only.

Operation Notes

  • The target prompt, image, and return files were checked as missing before writing.
  • The child-owned brief and packet existed before this run and are read-only for this worker.
  • The prompt uses the gpt-image-2 Skill as Advisor / prompt-engineering workflow.
  • Actual generation is through host built-in image_gen using the visible Image 1 and Image 2 references.
  • Garden/API execution is not expected for this project-bound run.

Advisor / Host-Native Record

Mode check command:

node C:\Users\wangq\.agents\skills\gpt-image-2\scripts\check-mode.js --json

Mode 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。"
}

Skill/template files read:

  • 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

Template adaptation:

  • concept-scene.md was used for foreground / midground / background / lighting structure, but the rendered prompt is constrained to photorealistic cinematic documentary road-film still, not digital concept art, poster, or illustration.

Garden/API expected: no.

Source files actually read for this worker:

  • AGENTS.md
  • docs/workflows/slide-task-orchestration.md
  • docs/workflows/slide-task-templates.md
  • projects/2026-06-23-在路上/slides/s07/s07-v2-brief.md
  • projects/2026-06-23-在路上/tasks/2026-06-24-character-consistency/session-06/packets/s07-v2-slide-agent-packet.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/visual-system.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
  • C:\Users\wangq\.codex\skills\.system\imagegen\SKILL.md

Final Prompt

Create a 16:9 photorealistic cinematic documentary road-film still.

Use Image 1 as the primary identity/body continuity reference for the tiny traveler: preserve the walking-away silhouette, long dark road-worn coat, old backpack shape, compact shoulder line, worn boots, and non-heroic posture. Use Image 2 as the supporting walking side-posture reference: preserve the natural stride, coat length and fall, backpack placement, and road-worn body profile. The traveler may be seen from a back or slight side-back angle, with no face emphasis.

Scene: a large cold muddy wilderness road in early morning or pre-dawn haze. The world appears before the person. A long worn road with wet ruts, frost, stones, sparse dead grass, dark puddles, rough earth, and low muddy ground runs through a bleak open landscape. Wind and airborne grit press sideways across the road and against the coat, creating physical pressure without becoming a fantasy storm.

Composition: environment first, road second, traveler third. Wide landscape-led road-film frame, not a portrait. The traveler is tiny, about 1/12 to 1/8 of the frame, placed off-center along the road line. Keep the road long enough to imply continuing movement into pressure. The first read should be cold wilderness, muddy road, and wind/sand pressure; only then should the viewer notice the small traveler as a moving coordinate.

Lighting and palette: low-saturation cold gray, earth brown, dark khaki, coal black, wet dark blue-gray. Cold morning haze or pre-dawn light, overcast and restrained. Documentary mud texture, gritty air, worn fabric, dirty wet road surface, old boots, and weathered backpack material. No broad sunrise, no golden scenic sunset, no clean tourism light.

Style: real documentary texture with restrained poetic realism, cinematic road-film still, natural lens perspective, believable scale, subtle grain, physical mud and wind. Landscape-led, quiet, pressured, unfinished, still walking.

Must keep: tiny secondary traveler; coat/backpack/walking silhouette consistent with Image 1 and Image 2; no face emphasis; no centered heroic backshot; no fashion or travel-ad pose; no vehicle or animal; no readable text.

Negative / Avoid

Avoid travel advertising, scenic postcard road, motivational poster, heroic centered stance, fashion photography, clean inspirational highway, tourism landscape, fantasy sandstorm, sci-fi, cyberpunk, neon, UI, code, logos, labels, watermarks, animals, vehicles, group reunion, victory tone, broad sunrise, broad sunset, high saturation, clean commercial styling, portrait MV, large face, glamour lighting.

Output Intent

  • Output file: projects/2026-06-23-在路上/slides/s07/s07-v2-image.png.
  • Intended result: S-07 Section A pressure shot, where wind/sand and muddy wilderness dominate and the traveler remains a tiny, consistent, road-worn coordinate.