nexu-io/openmotion / skills/openmotion-media
openmotion-media
VerifiedGenerate, poll and inspect individual image, video, speech or music assets through openmotion_assets_use after openmotion_media_use discovery (configured Atlas Cloud, fal.ai, WaveSpeedAI or OpenDesign Cloud BYOK providers, or a signed-in OpenDesign account), either when delegated by an OpenMotion film workflow or for a standalone asset-generation request. Native-only tasks with no new speech or music stay outside generation unless explicitly requested. Not a whole-film entry point: creating, recreating, templating, editing or exporting a whole video belongs to the OpenMotion video entries.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 6.4 KB |
| agents/openai.yaml | config | 359 B |
| references/inspection.md | markdown | 5.5 KB |
| references/jobs.md | markdown | 8.5 KB |
| references/planning.md | markdown | 3.6 KB |
| references/providers/account.md | markdown | 8.0 KB |
| references/providers/atlascloud.md | markdown | 7.4 KB |
| references/providers/codex.md | markdown | 1.6 KB |
| references/providers/compatible.md | markdown | 6.4 KB |
| references/providers/fal.md | markdown | 8.3 KB |
| references/providers/opendesign.md | markdown | 14.3 KB |
| references/providers/wavespeed.md | markdown | 6.4 KB |
| references/setup.md | markdown | 3.1 KB |
references/inspection.md
View on GitHub# Inspect and adopt media
Read when inspecting completed candidates, adopting selected versions or handing back results.
## Inspect a fixed version, then adopt it
Only `state: "completed"` and real `libraryAssets` confirm saved library candidates. Pin each chosen `{assetId, versionId}`; `get` returns its metadata, including measured dimensions/duration. `read` through `openmotion_assets_use` returns bounded image previews, up to four requested video frames or a mono 16 kHz audio excerpt (maximum 15 seconds). An audio excerpt reaches the active model as audio input: request it only when the active model and connection document audio input support. A rejected audio input can fail the whole model turn rather than return a tool error, so when support is absent or unknown, check speech and music with the local measurements below and any provider timing metadata instead. These previews do not prove full-video playback or complete listening; record checks unavailable in the host as unverified.
1. **Inspect before adopting**: use the exact library version for visual/audio checks, rather than the asset's changing current-version pointer. Reuse already approved assets. A standalone request can deliver its pinned version without creating a film project.
2. **Import only selected candidates into the film**: call the asset Tool with `action: "import_to_project"`, explicit `projectId`, stable `requestId`, `assetId` and `versionId`. Check per-item results and returned project-relative paths; import records a managed version but does not place media on the timeline. Then `openmotion_get_project({ projectId })` and use its refreshed `revision` / `currentVersionId` for editing. Probe the actual adopted file with `ffprobe` or available local media tools for width, height, real duration, frame rate and audio streams before timeline placement. Avoid concurrent project edits during adoption. Old `media_use` files are already in the project and must not be imported again.
3. **Inspect visual and audio usability**:
- **Images**: inspect decoded frames at intended crop and scale for subject fidelity, clean edges, artifacting and text-safe space. When an image will serve as an image-to-video first frame, also verify before spending on video that its aspect ratio fits the output canvas, its key details match the anchor, the subject starts in the intended beginning pose (not an already-finished state) and every moving part has full clearance inside the frame.
- **Video**: inspect sampled frames across the usable motion window for identity stability, unintended scene cuts, crop safety and embedded audio tracks. When several clips feature the same recurring subject, inspect representative start, mid and end frames side-by-side against the anchor and adjacent cut frames before adopting. When a generated video includes native or ambient audio, explicitly mute (`muted` / `volume = 0`) or mix that track in the composition so it does not collide with narration or music.
- **Speech and music**: verify spoken words, pronunciation, pacing and clean phrase endings; use `action: "get"` (`assetId` + `versionId`) plus `ffprobe` for container duration and `ffmpeg -af silencedetect` (or waveform inspection) to measure leading/trailing silence and actual spoken phrase end so composition cues, cuts and captions align to real speech rather than clipped endings or unintended black frames. Do not call `action: "read"` on audio unless the active host/model is confirmed to accept MCP `audio` input blocks (text-vision Codex models reject `input_audio` with HTTP 400 `Audio input is not available`); when transcript verification is needed without host audio input, run a configured transcription model or inspect metadata plus `silencedetect`. Currently only `opendesign` populates provider speech timing metadata (`alignment` or `sentences`) in `job.result` (`account` and other providers do not); read it from `status`, or pass both `assetId` and `versionId` to `action: "get"` to read top-level `version.result`, and record those timing intervals alongside the pinned version for downstream subtitle placement. If direct listening or visual inspection is unavailable in the host, record that check as unverified rather than passed.
## Hand back to the calling workflow
If setup interrupted the task, re-run provider inspection after confirmation and resume the original brief and selected assets; do not regenerate completed versions merely because account access changed.
In the delivery reply, name every part that is not generated as planned (for example local TTS, code-synthesized music or user/retrieved media) so the user knows what they received. For a standalone asset request, report the pinned asset/version identity and available file location, provider/model/job identity, measured dimensions or duration, and any quality or review limitations.
Inside a film workflow, pass back the pinned `{assetId, versionId}` together with measured speech windows and any `job.result` timing metadata, then continue with `openmotion_edit_project` using the refreshed `expectedRevision` and `expectedVersionId`. Generation produces candidate media, not a finished film: keep shot selection, native typography/layout, timeline choreography, Studio preview handoff and final audiovisual review in the active creative or production workflow. Generated assets never relax that workflow's own checks: segments using them pass the active workflow's applicable visual, text and measurement checks. The active workflow's delivery guidance defines which measurements gate generated shots.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| planning-and-task-breakdownaddyosmani/agent-skillsBreaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| screenpipethedotmack/claude-memSearch screen and audio history recorded by Screenpipe and optionally save selected context into claude-mem. Use when the user asks to recall desktop activity, a meeting, or something seen in another app through Screenpipe. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| contact-widgetnexu-io/open-designSelf-contained floating chat widget with welcome screen, social links, meeting button, and message input. Single HTML file, zero dependencies. | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |
| meeting-notesnexu-io/open-designMeeting notes page — title bar with attendees, agenda checklist, decisions block, action items table with owners + dates, and a "next meeting" footer. Use when the brief mentions "meeting notes", "minutes", "1:1 notes", "all-hands recap", or "会议纪要". | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |