Blackboard One-Shot Video
Paste the following prompt into your AI chat to install this skill:
Please follow https://skillhub.cn/install/skillhub.md to install @user_87b8e34f/blackboard-one-shot.
About this skill
Problem
When blackboard photos already exist in shoot order, a teacher may only need one independent short clip per photo, not a stitched montage. The hard part is that photos may not contain explicit formulas, sequence numbers, or narration; the skill must not invent missing text. Every remote generation is also billed, so scope, price, call count, and task identity must be clear before submission.
How it works
- Hard inputs: at least one accessible blackboard photo, arranged in shoot order; teacher facts for each photo, including lesson, board lines, and step or sequence number. Photos are not a source for missing text; ask only for missing required inputs.
- Free plan first: inspect MIME, dimensions, aspect ratio, byte size, and alpha channel; upload only through the bundled client. Then output a one-photo, one-shot plan. Photos with missing facts remain in the plan, but planning is not approval.
- Billing confirmation: query
beatra.models.list, confirmimage_to_videosupports a first-frame image withoutdriving_audio; show what will be done, credits, call count, identity, stop condition, and low-balance handling. If balance is insufficient, keep the official recharge link and do not retry until the user confirms. - Generate per photo: after confirmation, call
beatra.videos.animateonce per accepted photo with a new opaqueclient_request_id, pollbeatra.tasks.getto terminal state, deliver in shoot order, and check identity, motion, and detail drift.
Boundaries
Suitable for teacher blackboard photos to single-shot clips; use wrong-item-talking for photo plus narration, assembly-one-step-clip for assembly steps, image-to-motion for generic stills, and tool-demo-clip for enterprise tool photos. Do not fill missing board text from public posts, and do not invent concat, editing tools, or cross-photo splicing; changed payloads are new work and require a new task ID.
Use Cases
- Teachers who already have shoot-ordered blackboard photos need to generate one 2-15s single-shot clip per photo without stitching, splicing, or inventing missing board text.
- Course leads need a free annotated one-photo, one-shot plan with model constraints and billing quote before submitting any Beatra image-to-video generation task.
- Educational operations staff need to verify per-photo task IDs, polling terminal states, actual dimensions, usage, and billed credits after generation.
- Instructional designers need to deliver clips in original shoot order and avoid adding concat, editing tools, or public-post text for missing board facts.
Best For
- K-12 teachers who need to turn already-shoot blackboard photos into one short single-shot clip per photo without stitching.
- Course production leads who need to deliver teaching clips in original shoot order without splicing.
- Educational operations staff who need to verify model, credit, and call-count checks before submitting generation.
- Teaching assistants who need to avoid inventing missing board text and keep the original photo order.
Related Skills
Offers mock interviews, question-bank drills, and single-question polish, using STAR, recruiting timelines, and group interviews to help engineers refine answers from real experience.
Use Bloom's taxonomy to structure concept learning from recall to creation and generate lightweight Markdown notes for Obsidian.
A goal-specific interview prep coach for JD decoding, story framing, mock interviews, and follow-up review.
Designs sales team training programs end to end, covering needs analysis, ABCD objectives, ADDIE or 6D course structure, materials, Kirkpatrick evaluation, and timelines.