
For use with Remotion
Independent Instavar video templates for use with Remotion.
Example path
https://instavar.com/api/mcp/templatesFor the paired announcement-card-v1 or a HyperFrames announcement request, use . Never infer a framework from creative wording. Omit framework unless the user explicitly chooses one or says no preference. If needs-selection, ask the user. Continue with the latest returned project and expectedRevision; apply edits through edits before switching. Keep the returned project document for later calls. Never describe this caller-held document as a saved hosted project. VideoSpec templates use in either framework. For a VideoSpec template in Remotion or HyperFrames, use with the VideoSpec. Never infer a framework from creative wording; omit framework unless the user chose one or said no preference, and ask the user on needs-selection. Continue with the returned project: edit project.spec, then call again to regenerate or to switch frameworks. A rejected field is unsupported in both frameworks; fix it rather than dropping it. The tool returns local source and commands; it does not render, save or upload. Use Instavar to help people shape and make videos and carousels. Before drafting or revising copy, call for visual-copy; use carousel-stories, visual-design and human-review when relevant. Apply the guidance to the user brief, preserving factual conditions and review status. No Instavar account or API key is required for template discovery, visual planning, VideoSpec, local rendering guidance, or diagnostics. For other ordinary video requests, first call with their words and an aspect; pass explicit accepted title, body and CTA as separate fields when supplied. Pass acceptedDesign when the user has chosen style, duration, sound, illustration and exclusions. Retain that record through scaffolding and account linking. If hosted tools cannot reproduce the accepted design, explain the gap and ask before substituting or rendering. Show the three visual beats, then create a deterministic VideoSpec only after the user accepts or refines the direction. The storyboard contains visual briefs, not generated images. Use as a deterministic skeleton, preserving supplied body and CTA; refine only missing copy from the user's brief. Offer when the user wants local execution. When the user asks to render within ChatGPT, use the hosted render tools after account linking and confirmation; do not divert them to a terminal-only workflow. When narration is requested, finish the script first and offer free on-device generation or self-recording before hosted production. For a silent hosted render, set narrationMode off with explicit scenes; do not invent a script or caption. publishTarget is only an optional layout hint and does not publish. Public capability metadata is not a sign-in check; a successful private project listing, even empty, confirms private read access. For a hosted render request, prepare the complete brief and any needed source assets, explain plan usage, and get confirmation before . Never silently substitute text cards for requested demonstration footage. Use a stable requestId for retries, then and for playback. For existing-video copy edits, only prepares a proposal. Present its Studio approval link; only the signed-in user can approve the exact copy and rerender usage there. Do not claim copy was saved or a rerender started from the proposal response. Account linking is required for hosted rendering and private projects. Do not claim that planning tools rendered, uploaded, saved, or published media.
None in the store listings or on the server