← WidemileCONTENT HISTORY

Update to Widemile

Snapshot Oct 24, 2026 · 13:07 UTC · version 1.1.0

Collection source: downloaded plugin package.

WHAT CHANGED

First saved snapshot

No earlier snapshot is available to establish a change.

Save Widemile and check future changes in Following.

Up to 3 plugins free · no card required.

Compare saved observations

Download comparison JSON
Full technical diff · 0 changed fields
Full snapshot data
{
  "description": "Create one cinematic bullet-time VFX shot from one supplied source image or video: normal motion decelerates into a precise time freeze, the camera completes one sweeping 360-degree orbit around the frozen subject and scene, then time resumes naturally; use for freeze time, bullet time, frozen moments, suspended action, freeze-frame orbit, tempo congelato, bullet time, or tiempo congelado, not for a static image, an ordinary camera orbit, a camera spin, a hard cut, or a generic slow-motion shot.",
  "included_files": [],
  "name": "vfx-360-in-time",
  "skill_md_contents": "---\nname: vfx-360-in-time\ndescription: \"Create one cinematic bullet-time VFX shot from one supplied source image or video: normal motion decelerates into a precise time freeze, the camera completes one sweeping 360-degree orbit around the frozen subject and scene, then time resumes naturally; use for freeze time, bullet time, frozen moments, suspended action, freeze-frame orbit, tempo congelato, bullet time, or tiempo congelado, not for a static image, an ordinary camera orbit, a camera spin, a hard cut, or a generic slow-motion shot.\"\n---\n\n# 360 in Time VFX\n\n## Tool requirements\n\n- **Core evidence and source binding:** `prostudio_brief_get`, `prostudio_workspace_get`, `prostudio_references_list`, `prostudio_references_resolve`, `get_prostudio_media_catalog`, and `inspect_prostudio_media`.\n- **Final 360 in Time clip:** `prostudio_models_list`, `prostudio_generation_prepare`, `prostudio_video_generate`, `prostudio_job_status`, and `prostudio_generation_get`.\n- This declaration does not activate tools. Use only the tools required by the selected workflow. If a required tool is absent from the current scoped catalog and `load_prostudio_tools` is available, call it with the missing outcome before claiming the capability is unavailable. If retrieval still exposes no matching tool, report that limitation. A model-catalog lookup alone is never completion for a requested 360 in Time shot.\n\nCreate one high-end cinematic VFX clip in which the supplied scene first behaves normally, time decelerates into a motivated freeze, the camera travels around the frozen moment through one complete 360-degree orbit, and the original action then resumes with a clean causal release. This is a **bullet-time time stop**, not a camera-only spin or a still-image hold. The frozen moment must remain spatially alive: the camera changes viewpoint, foreground and background layers create parallax, depth and occlusion stay coherent, and every suspended subject and object remains locked to the same instant until the release.\n\nUse semantic skill selection only. Do not add deterministic routing, keyword matching, catalog changes, or unrelated UI changes as part of this workflow.\n\n## Source policy\n\n- The workflow accepts exactly one user-selected base asset: either one image or one video. Treat that asset as the reference authority for the shot rather than asking for a second plate by default.\n- Load `media-inspection` and inspect the selected source before planning. Filenames, thumbnails, prompts, and summaries are not evidence that the frame, identity, motion, camera, lighting, or resolution is suitable.\n- Load `reference-management` and bind the selected source explicitly as the Freeze origin: subject identity, scene geography, action, camera grammar, time of day, light direction, wardrobe, props, and visible text all remain protected unless the user requests a change.\n- Do not require a clean plate, depth map, object mask, motion-control pass, or separate orbit reference when the selected live video model can derive the shot from the supplied source.\n- If the user supplies a video, preserve the reference video's performance and visual identity. Keep its subject, action, framing, camera language, environment, exposure, lighting, and temporal cadence before and after the effect. Apply the time freeze only to the most meaningful, physically readable moment; do not remake the whole video as an unrelated scene.\n- If the user supplies an image, reconstruct the complete scene from that reference. Infer coherent foreground, subject, midground, background, ground contact, depth, occlusion, and lighting so the subject and environment can animate before and after the freeze and support a full camera orbit. Do not treat the image as a flat card or invent a different location, identity, wardrobe, or visual premise.\n- For an image source, preserve the reference composition and visual facts while deriving only the missing temporal behavior: a short lead-in action, the freeze trigger, the orbit, and the return to motion. Keep all newly animated details subordinate to the supplied reference.\n- If the source has several plausible subjects or no clear moment to freeze, ask one concise question that identifies the intended anchor or trigger. Otherwise infer non-critical timing from the action and continue. Never require a Brief to run Freeze; a Brief is optional context.\n\n## Define the freeze moment\n\n- Choose a visible, motivated trigger already supported by the source: a step, turn, impact, gesture, falling object, splash, hair movement, thrown prop, reaction, or another readable action beat. Do not invent a spectacular event merely to justify the effect.\n- The trigger must be causal: normal action leads into a short deceleration, all relevant motion settles at one exact instant, the orbit exposes that same instant from multiple angles, and the original action continues from the same physical state.\n- Freeze the whole visible scene at the chosen instant, not only the main subject. People, props, droplets, dust, smoke, hair, clothing, vehicles, foliage, reflections, and other moving elements hold their positions and shapes. Lighting may remain naturally alive only when that is consistent with the source; it must not create a new event during the freeze.\n- Preserve contact and spatial relationships throughout the orbit: planted feet stay planted, hands remain attached to objects, suspended objects retain their trajectory context, shadows remain attached to their causes, and foreground occlusions reveal real depth rather than cutout layers.\n- The main subject remains the visual anchor, but the orbit must reveal the surrounding scene. Do not lock the subject to a flat frontal portrait or rotate only the image around its center.\n\n## Execute the bullet-time grammar\n\nAuthor the final result as one continuous shot with these five causal phases. Rebalance the time ranges to the selected duration, but do not remove the full orbit or reorder the phases:\n\n1. **Normal baseline:** the source scene is alive and faithful to the selected video, or the image-derived scene begins with restrained, believable motion. Establish the subject, action, camera height, lens character, and spatial anchors.\n2. **Trigger and deceleration:** the subject reaches the chosen action beat. Motion eases down over a short ramp; hair, fabric, loose objects, dust, water, and camera inertia decelerate together. The effect must feel like time is stopping, not like a playback-speed preset was applied.\n3. **Time freeze and 360 orbit:** all scene action is suspended at the trigger instant. The camera completes exactly one full 360-degree orbit around the subject or the defined action center, returning to its starting azimuth after viewing the front, side, rear, and opposite side as the available framing allows. The orbit is a real change of viewpoint with parallax, depth, perspective, foreground occlusion, and coherent lighting; it is not a lens-axis spin, digital rotation, zoom, or loop.\n4. **Release:** after the orbit reaches its completed 360-degree return, time resumes from the same poses, object locations, foot contacts, and camera orientation. Use a brief, natural snap or smooth ramp according to the source's energy; never teleport the subject or reset the environment.\n5. **Normal resolution:** the original action continues for long enough to read. Preserve identity, screen direction, performance, camera momentum, exposure, and environment continuity. End on the natural consequence of the action, not on a frozen frame.\n\nThe default timing for an 8-second result is approximately 0.0–2.0 seconds normal baseline, 2.0–2.7 seconds trigger and deceleration, 2.7–5.8 seconds frozen 360-degree orbit, 5.8–6.4 seconds release, and 6.4–8.0 seconds normal resolution. Use a longer freeze only when the orbit needs room to read; do not compress the full 360 into an unreadable whip.\n\n## Direct the camera and scene\n\n- Define a stable orbit center at the subject's torso, face, or the clear action focal point. The camera should travel on a controlled circular or slightly elevated 3D path around that center and return to the original azimuth, height relationship, and subject scale at the end of the orbit.\n- Use real camera translation and changing perspective. Foreground objects should pass through the frame at different rates, the background should exhibit depth-dependent parallax, and occlusions should update as the camera moves. Keep the subject anatomically stable while the viewpoint changes.\n- A strong default is a 28–40 mm full-frame-equivalent lens, a smooth stabilized dolly/orbit at subject height, a modest rise or fall only when supported by the scene, shallow-to-moderate depth of field, natural motion blur during the approach and release, and very restrained blur during the frozen orbit. Avoid fisheye distortion and stretched faces.\n- Keep the camera's horizon level and rotation axis intentional. A 360-degree orbit around the subject is required; a 360-degree roll around the lens axis, a spinning background, or a camera yaw with no translation does not satisfy the skill.\n- Preserve the source video's camera path before and after the effect. During the freeze, extend the established visual language into the orbit rather than replacing a handheld, tracking, or locked composition with an unrelated drone move.\n- For an image source, animate the entire reconstructed scene with coherent camera parallax. Add only source-supported motion before and after the freeze: breathing, a step, a glance, wind, practical lights, passing atmosphere, foliage, traffic, or reflections when they belong to the reference.\n- Keep the subject readable through changing backlight, foreground crossings, and rear-view angles. Use motivated exposure continuity, atmospheric depth, contact shadows, reflections, and rim separation instead of artificial cutout edges.\n- Aim for an ARRI ALEXA 35 or LF-style cinematic image when the request is photoreal: natural skin tones, controlled highlight roll-off, detailed shadows, restrained film grain, realistic motion blur, and a coherent LogC-style grade. Treat this as capture intent only; never claim a physical camera, lens, codec, or high-frame-rate capture unless the tool reports it.\n\n## Audio and temporal continuity\n\n- Preserve the source video's audio only when it remains appropriate and the selected generation mode supports it. Otherwise define a clean audio treatment rather than silently carrying unrelated audio.\n- A convincing default is normal ambience and performance audio, a brief filtered deceleration or suction as time stops, a restrained orbit bed or suspended ambience during the freeze, a precise release transient, then the original ambience returning at normal cadence. Do not let audio imply that the camera is spinning while the world is frozen.\n- For a still image, do not claim source audio exists. Choose silence, generated atmosphere, or a minimal cinematic cue only if supported by the request and live model. Audio must not replace the visual proof of the freeze.\n- Maintain continuity of screen direction, light direction, subject scale, wardrobe, facial identity, object ownership, and scene geography through all five phases. The release must continue from the exact frozen state rather than morphing into a new pose.\n\n## Prompt strategy\n\n- Load `generation-planning`, inspect the live model catalog, and select a currently available reference-compatible video mode. Do not hardcode a provider or promise deterministic identity, exact physical simulation, or unsupported image/video conditioning.\n- Prepare exactly one final video generation and execute exactly one `prostudio_video_generate` action. Do not render the baseline, freeze orbit, and release as separate clips and do not use the Editor to assemble them.\n- Write one spatially explicit English prompt. Name the bound source asset, source type, identity and environment locks, freeze trigger, orbit center, exact 360-degree path, depth/parallax behavior, timing, camera package, aspect ratio, frame rate, motion-blur intent, light direction, and audio behavior.\n- Use this structure and replace every bracketed field with concrete information:\n\n  ```text\n  ONE CONTINUOUS HIGH-END CINEMATIC BULLET-TIME FREEZE VFX SHOT, one supplied [IMAGE/VIDEO] source, one finished clip, no montage.\n  SOURCE LOCK: [exact bound source asset and source type]; preserve [identity, wardrobe, action, environment, framing, camera language, lighting, time of day, visible text, and color intent].\n  ACTION ANCHOR: [subject and readable trigger already present or supported by the source], centered at [orbit center].\n  PHASE 1 [start–end] NORMAL BASELINE: [faithful source action or image-derived scene animation], natural camera motion, stable screen direction, and readable spatial anchors.\n  PHASE 2 [start–end] TRIGGER AND DECELERATION: [exact action beat] slows with a smooth motivated ramp; [hair, fabric, loose objects, particles, reflections, and camera inertia] decelerate together into one exact frozen instant.\n  PHASE 3 [start–end] TIME FREEZE / FULL 360 ORBIT: time is completely stopped at [frozen instant]. The camera physically travels around [orbit center] through exactly one complete 360-degree orbit, with real translation, changing perspective, foreground occlusion, layered parallax, stable depth, coherent light and shadows, and the subject's anatomy and all suspended objects locked in place. Return to the starting azimuth and subject scale. This is not a camera roll, lens-axis spin, digital rotation, zoom, loop, or camera-only move.\n  PHASE 4 [start–end] RELEASE: after the completed 360-degree return, time resumes from the identical poses, contacts, object positions, light direction, and camera orientation with [brief snap or smooth ramp].\n  PHASE 5 [start–end] NORMAL RESOLUTION: [original action continues and resolves], preserve identity, performance, camera momentum, environment geography, exposure, and final composition.\n  CAMERA: [capture intent, focal length, height, orbit radius, focus, depth of field, frame rate, shutter/motion blur, stabilized movement, level horizon, aspect ratio].\n  LIGHT, GRADE, AND AUDIO: [source-matched light and grade]; [normal ambience, suspended freeze treatment, release cue, and return to normal audio or silence].\n  CONTINUITY: one source authority, one identity, one scene, one trigger, one exact 360-degree orbit, one freeze instant, and one causal return to motion.\n  ```\n- For a video source, state the exact source action and the small temporal window where the VFX is applied. Explicitly say that the unaffected lead-in and tail retain the reference video's performance and visual signature.\n- For an image source, state the complete reconstructed scene, the depth order of foreground/subject/midground/background, the minimal lead-in and tail animation, and the rule that the image becomes alive rather than remaining a frozen card.\n- Put positive behavior before constraints. Then explicitly forbid a static photo hold, frozen cutout, image-plane rotation, lens-axis spin, orbit with no parallax, incomplete orbit, more than one orbit, camera roll, arbitrary yaw, zoom-only move, hard cut, dissolve, scene replacement, identity drift, duplicated subject, extra limbs, broken object contacts, floating feet, melting geometry, exposure flash, temporal loop, unrelated new action, portal, tunnel, smoke wipe, decorative particles, and a release that starts from a different pose.\n\n## Prepare, generate, and inspect\n\n- Before generation, verify one exact source asset is bound and visually suitable; source type is known; the trigger, orbit center, freeze instant, full 360 path, camera return, and release are explicit; image-derived depth and animation are defined; video fidelity boundaries are defined; and the live model supports the selected binding, duration, aspect ratio, and audio behavior.\n- Prepare the exact generation and report the returned model, source type, bound asset, duration, aspect ratio, audio behavior, and credit charge, then stop for the host's required confirmation. After approval, execute the prepared action without changing its arguments.\n- Poll the durable job and inspect the completed video before calling the Freeze result successful. A completed job is not proof that time actually froze or that the orbit completed a full 360 degrees.\n- Inspect the lead-in, the trigger ramp, the whole frozen orbit, the return to the starting azimuth, the release, and the final action. Check identity, anatomy, object contacts, depth, parallax, occlusion, shadow attachment, exposure, focus, motion blur, audio continuity, and whether the source video's unaffected portions remain faithful.\n\n## Preflight checklist\n\nBefore generation, verify that:\n\n- exactly one image or one video is the source authority;\n- a visible, motivated trigger and one exact frozen instant are defined;\n- the camera completes one real 360-degree orbit around a named subject or action center and returns to its starting azimuth;\n- the frozen scene remains spatially alive through translation, parallax, perspective, occlusion, and coherent depth;\n- every relevant subject, prop, particle, shadow, reflection, and environmental element is suspended consistently;\n- a video source keeps its reference performance, camera language, framing, and environment outside the localized Freeze window;\n- an image source is reconstructed as a complete, animated scene with layered depth and a readable lead-in and tail;\n- the five phases fit the requested duration and the orbit is not rushed into an unreadable whip;\n- the prompt contains one final generation with no placeholders, separate phase renders, or Editor assembly;\n- audio behavior is explicit and does not contradict the visual time stop;\n- the live model supports the exact source binding and requested output settings.\n\n## Review and revision criteria\n\n- **Pass:** normal motion visibly leads into a motivated time stop, every relevant element freezes at the same instant, the camera completes one full 360-degree physical orbit with parallax and changed viewpoint, and motion resumes from the exact frozen state.\n- **Pass:** the subject remains recognizable and anatomically stable; feet, hands, props, shadows, reflections, lighting, wardrobe, screen direction, and scene geography remain coherent throughout the orbit and release.\n- **Pass:** a video result preserves the reference video's visual identity and applies the effect only where the action benefits from it; an image result reconstructs a complete scene that can move before and after the freeze without looking like a flat 2.5D card.\n- **Pass:** camera motion is smooth, motivated, and cinematic, with controlled lens perspective, depth of field, exposure, focus, and motion blur; the horizon and orbit axis remain intentional.\n- **Fail:** the result is only slow motion, a still frame, an image-plane rotation, a lens-axis spin, a digital zoom, a camera roll, or a camera move with no depth/parallax.\n- **Fail:** the orbit stops short of 360 degrees, loops more than once, changes center, returns to the wrong azimuth, or hides the freeze behind a cut, dissolve, wipe, portal, cloud, smoke, or abstract transition.\n- **Fail:** the subject or environment morphs, duplicates, melts, changes wardrobe, loses identity, breaks object contacts, floats, slides, or resumes from a different pose; shadows, reflections, particles, or audio continue moving independently without a source-supported reason.\n- **Fail:** a video source is remade wholesale instead of preserving its action and visual signature, or an image source remains a frozen card with only a rotating camera.\n- For a localized failure, revise only the trigger timing, freeze window, orbit center/radius, 360 path, depth/parallax cue, source-fidelity constraint, audio treatment, or release continuity, then regenerate the complete single clip. Do not patch the result with a separate orbit or an Editor composite.\n"
}

SHA-256 of public snapshot: dfd5475d95d2e10ce298d70fe1e8d91d015554738437b283a1249aeacfd8e737