Skill instructions
ad-multiplier2.17 KB
View saved version →
---
name: ad-multiplier
description: "Produce independent edited variations of a supplied video ad in Sequencer, replacing people, garments, products, locations, or targeted details while preserving the original performance, cuts, timing, and audio. Use for an existing ad, not new campaign concepts or videos from scratch."
---
# Ad Multiplier
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `ad-multiplier`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Original video ad. Ordered changes and corresponding references.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
ad-variant-matrix2.18 KB
View saved version →
---
name: ad-variant-matrix
description: "Plan and produce a bounded set of advertising creatives from one brief in Sequencer, with named test variables, reusable assets and an export manifest. Use for new campaign variants and controlled creative tests; route edits preserving an existing ad's performance and timing to Ad Multiplier."
---
# Ad Variant Matrix
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `ad-variant-matrix`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Product or offer, audience and campaign objective.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
apartment-listing2.03 KB
View saved version →
---
name: apartment-listing
description: "Turn a real apartment listing into a polished video using verified property photos and facts. Run this workflow with Sequencer."
---
# Apartment Listing
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `apartment-listing`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: One listing URL or listing facts and verified property photos.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
architectural-walkthrough2.07 KB
View saved version →
---
name: architectural-walkthrough
description: "Create architectural walkthrough videos, interior tours or property flythroughs from supplied plans, renders or reference photos. Run this workflow with Sequencer."
---
# Architectural Walkthrough
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `architectural-walkthrough`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Architectural references and the spaces to show.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
article-to-audio1.99 KB
View saved version →
---
name: article-to-audio
description: "Turn one article into a faithful, approved, verified spoken-audio file. Run this workflow with Sequencer."
---
# Article to Audio
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `article-to-audio`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: One public article URL or supplied article text.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
article-to-podcast2.11 KB
View saved version →
---
name: article-to-podcast
description: "Turn one article into an accurate, approved single-narrator audio rendition. Run this workflow with Sequencer."
---
# Article to Podcast
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `article-to-podcast`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: One public article URL or supplied article text. Adaptation or read-as-written treatment, language, voice direction and target duration when materially unresolved.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
article-to-video2.01 KB
View saved version →
---
name: article-to-video
description: "Turn one article into a source-grounded, narrated, finished video. Run this workflow with Sequencer."
---
# Article to Video
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `article-to-video`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: One article URL or supplied article text, with requested video treatment.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
brandkit2.03 KB
View saved version →
---
name: brandkit
description: "Create, extend or apply brand guidelines, palettes, typography and a coordinated visual identity across deliverables. Run this workflow with Sequencer."
---
# Brandkit
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `brandkit`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Brand name, supplied identity assets and requested applications.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
character-identity2.12 KB
View saved version →
---
name: character-identity
description: "Create and reuse a consistent character reference set in Sequencer for images and video. Use for recurring people, mascots or fictional characters; this is reference-based continuity, not identity-model training or live-avatar setup."
---
# Character Identity
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `character-identity`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: A character brief or supplied identity references.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
clipping2.15 KB
View saved version →
---
name: clipping
description: "Turn supplied long-form video or podcast audio into finished short clips in Sequencer. The agent selects moments, edits source footage, reframes, captions, reviews and exports. Use for extracting faithful highlights from existing media, not generating replacement performances or new videos from a brief."
---
# Clipping
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `clipping`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Accessible source video or podcast audio.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
create-avatar2.03 KB
View saved version →
---
name: create-avatar
description: "Build a live character with guided appearance, personality, voice, and framing choices. Run this workflow with Sequencer."
---
# Create Avatar
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `create-avatar`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Character purpose, personality and a portrait or approved portrait-generation plan.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
create-game1.99 KB
View saved version →
---
name: create-game
description: "Plan and build an editable browser game saved in the workspace Games tab. Run this workflow with Sequencer."
---
# Create Game
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `create-game`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Game idea and what players should do. An accessible workspace.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
create-world2.01 KB
View saved version →
---
name: create-world
description: "Build a Worldstream with Genesis, custom interfaces, and interactive game logic. Run this workflow with Sequencer."
---
# Create Worldstream
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `create-world`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Worldstream concept or requested edits and an accessible workspace.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
creative-studio2.11 KB
View saved version →
---
name: creative-studio
description: "Generate or transform individual images, clips, speech, music or sound effects in Sequencer. Use for direct media requests; route complete ads, identity systems and multi-scene films to their dedicated workflows."
---
# Creative Studio
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `creative-studio`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Requested media type and a creative brief or source references.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
direct-believable-acting1.19 KB
View saved version →
---
name: direct-believable-acting
description: "Give characters playable tasks, attentive listening, and motivated reactions. Run this workflow with Sequencer."
---
# Direct Believable Acting
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `direct-believable-acting`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Expected inputs: A scene, character or performance to direct.
Return the requested writing or audit in the conversation. Workspace setup, generation and paid-production approval are required only for separately requested production work.
Referenced files: 1 — file references are metadata, not package contents.
faceless-youtube-video2.18 KB
View saved version →
---
name: faceless-youtube-video
description: "Create a narrated YouTube video or short explainer without an on-camera presenter. Use for educational explainers, history, science and voice-led illustrated stories. Honor short runtimes such as 15 seconds. Run this workflow with Sequencer."
---
# Faceless YouTube Video
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `faceless-youtube-video`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Topic or source material. Target viewer and requested runtime or format when unresolved.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
fashion-editorial2.16 KB
View saved version →
---
name: fashion-editorial
description: "Create editorial fashion photographs in Sequencer using a restrained, silhouette-led composition, tactile materials, compressed tones, and deliberate negative space. Use for fashion editorials and garment-led stills; not catalog packshots, virtual try-on accuracy, or fashion video."
---
# Fashion Photography
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `fashion-editorial`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Garment references and an editorial brief.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
film2.06 KB
View saved version →
---
name: film
description: "Build a cohesive multi-scene film from a concept, approved storyboard, and finished editable video. Run this workflow with Sequencer."
---
# Short Film
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `film`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Film idea or story brief. Audience, format and runtime when the request leaves a materially important choice unresolved.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
gallery-world2.03 KB
View saved version →
---
name: gallery-world
description: "Build a browsable gallery or arcade with collections and playable experience pages. Run this workflow with Sequencer."
---
# Gallery & Arcade World
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `gallery-world`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: A gallery or arcade brief and the real workspace experiences to display.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
hosted-world2.01 KB
View saved version →
---
name: hosted-world
description: "Build a complete JavaScript site with custom pages for your avatars, games, and Worldstreams. Run this workflow with Sequencer."
---
# Create World
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `hosted-world`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Hosted-site brief or an existing hosted world to edit.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
logo-animation2.01 KB
View saved version →
---
name: logo-animation
description: "Animate a supplied logo into a brand reveal, intro, outro or loop while preserving the approved mark. Run this workflow with Sequencer."
---
# Logo animation
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `logo-animation`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Approved logo asset and intended placement.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
marketplace-cards2.12 KB
View saved version →
---
name: marketplace-cards
description: "Produce product listing images, feature cards and A+ style modules in Sequencer from verified product references and marketplace requirements. Use for a listing set rather than generic lifestyle photography."
---
# Marketplace Cards
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `marketplace-cards`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Verified product references, product facts and target marketplace placements.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
motion2.13 KB
View saved version →
---
name: motion
description: "Create product and brand motion in Sequencer from a still, render, logo, or brief, including cinematic reveals and website product loops. Use for material, light, camera, and mark animation; not avatar explainers, character stories, or variations of an existing ad."
---
# Create Motion
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `motion`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Product, render, logo, still reference or motion brief.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
news-to-video2.1 KB
View saved version →
---
name: news-to-video
description: "Create a source-grounded news video with an approved script, captions and editable animated graphics. Run this workflow with Sequencer."
---
# News to Video
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `news-to-video`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: One public reporting URL or supplied article text. Format, length, narration choice, tone, and voice/language when narration is requested.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
pitch-trailer-animatic2.17 KB
View saved version →
---
name: pitch-trailer-animatic
description: "Turn a concept, treatment or series bible into a timed animatic, generated pitch trailer or supplied-footage sizzle reel in Sequencer. Use for communicating a story and visual direction to buyers or collaborators, with consistent references, planned timing and finished exports."
---
# Pitch Trailer / Animatic
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `pitch-trailer-animatic`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Story concept, treatment or series bible.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
product-photoshoot2.06 KB
View saved version →
---
name: product-photoshoot
description: "Create product photographs, studio shots, lifestyle images, hero images or coordinated product carousels from product references. Run this workflow with Sequencer."
---
# Product Photoshoot
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `product-photoshoot`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Product images or a product URL and the requested image set.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
reframe2.01 KB
View saved version →
---
name: reframe
description: "Expand existing videos to vertical, square or landscape formats while retaining full source duration and original audio. Run this workflow with Sequencer."
---
# Reframe videos
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `reframe`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Source video and requested output formats.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
remotion-motion-design2.01 KB
View saved version →
---
name: remotion-motion-design
description: "Create editable motion graphics with React animation layers and export a finished video. Run this workflow with Sequencer."
---
# Remotion Motion Design
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `remotion-motion-design`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Message, subject, or supplied copy.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
restaurant-menu-to-video2.1 KB
View saved version →
---
name: restaurant-menu-to-video
description: "Create a restaurant promotion from exact menu information and verified dish images. Run this workflow with Sequencer."
---
# Restaurant Menu to Video
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `restaurant-menu-to-video`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: A menu or restaurant URL, or supplied menu file/text. Verified dish photos and brand assets where specific dishes need to appear.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
retro-web-world2.02 KB
View saved version →
---
name: retro-web-world
description: "Build a vintage Web 2.0 world with glossy panels, badges and custom experience pages. Run this workflow with Sequencer."
---
# Retro Web World
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `retro-web-world`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: The intended visual era, subject and hosted-world experience lineup.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
script-to-storyboard2.03 KB
View saved version →
---
name: script-to-storyboard
description: "Turn a supplied script into an illustrated shot plan or storyboard before video production. Run this workflow with Sequencer."
---
# Script to storyboard
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `script-to-storyboard`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Script and any character, location or visual references.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
sequencer5.15 KB
View saved version →
---
name: sequencer
description: Create and edit images, videos and audio with Sequencer. Use for product ads and variations, architectural walkthroughs, short films, avatar explainers, social clips, article or webpage videos, storyboards, product photography, thumbnails and branded visuals. Animate stills, edit footage, expand aspect ratios, upscale media, remove backgrounds, convert video to HDR, generate voiceovers, music and sound effects, and assemble editable scenes with captions, transitions and motion graphics. Also use to browse or revise Sequencer projects, discover workflows, compare supported models, quote production costs and export finished videos.
---
# Sequencer
Use the Sequencer MCP tools for work on the user's Sequencer account and projects.
If the required tools are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
## Connection
- Use the hosted `sequencer` MCP server supplied by this plugin.
- If authorization is required, start the Sequencer sign-in flow and let the user approve access.
- Never ask the user to paste OAuth access tokens, Firebase tokens, API keys, or device session tokens into chat.
## Workflow
Preserve an explicitly chosen provider. Answer explanations without starting production. For a request to create or edit architectural walkthroughs, short films, avatar explainers, product motion, ad variations or other multi-step media, call `list_sequencer_skills`, match the user goal to the returned descriptions, and load the best match with `get_sequencer_skill`. Briefly explain the choice. The live catalog includes newly published workflows; proceed normally if none applies. For an explicitly named skill, discover its current ID and load it before planning or generation.
1. Establish context with `list_workspaces`, `list_edits`, or another narrow read tool.
2. Before timeline changes, call `get_edit_full` so current scenes, shots, media, audio, overlays, transitions, and exports are understood together.
3. Prefer narrow mutation tools such as `update_shot`, `update_shot_media_ref`, and `update_audio_track` over broad replacements.
4. Generate or upload media before attaching its returned media ID to a shot, asset, overlay, or audio track.
5. Preserve existing work. Do not clear fields, delete records, spend credits, or export unless the user's request calls for it.
6. For subjective changes, use review tools first and apply only recommendations that match the user's intent.
7. Report the workspace, edit, and meaningful results after mutations, including IDs or output URLs that help the user continue.
## Confirmations
- Treat generation, AI editing, review exports, voice changes, and final exports as credit- or resource-consuming actions. Call them only when the user's request clearly authorizes that outcome.
- Before deleting a scene, shot, asset, media reference, transition, or audio track, identify the exact target and obtain confirmation unless the user already explicitly named that deletion.
- Never make a project public, publish media externally, or share a private project link unless the user explicitly asks.
## Reliability
- Before choosing generation models, call `get_model_catalog` and follow its `modelSelectionGuidance` for the use case. Use the returned list to select an exact model ID and inspect supported inputs, durations, resolutions, audio options, and reference limits. Keep public model names unchanged and internal routing private.
- Before generation, call `quote_generation` with every billable setting. Show the model name, settings, USD total, timestamp, and whether the result is an estimate. Resolve missing inputs before presenting a total. Respect a budget the user already approved; otherwise ask for approval of the concrete plan and quote.
- Pass the approved `maxCostUsd` and a stable `idempotencyKey` to generation tools when suppo
Preview of the first 4,000 characters. Read full instructions →
Referenced files: 1 — file references are metadata, not package contents.
sequencer-video1.28 KB
View saved version →
---
name: sequencer-video
description: Create or edit videos including architectural walkthroughs, short films, avatar explainers, product animations, ad variations and webpage-to-video projects. Use to discover the Sequencer workflow for a video request spanning multiple steps.
---
# Sequencer video
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
Preserve an explicitly chosen provider. For a request to create or edit a video, call list_sequencer_skills, select the best matching workflow by description and required inputs, then load its exact ID with get_sequencer_skill before planning. Briefly explain the choice. If no workflow fits, proceed with available tools within the requested scope.
For explanations, answer the question without starting production. For planning requests, prepare only the plan. Resolve essential missing inputs before quoting. Obtain approval of the concrete production scope and cost before paid generation and preserve already approved choices.
Referenced files: 1 — file references are metadata, not package contents.
setup1.88 KB
View saved version →
---
name: setup
description: Guide a new Sequencer plugin user to connect their existing account and start a product ad, image animation, or editable short video.
---
Help the user get a useful first result with Sequencer.
1. Preserve their current request if they installed Sequencer during a conversation. Otherwise offer three starting tasks: make a product ad, animate an image, or plan a three-shot video. Ask at most one question at a time.
2. If account connection is required, use ChatGPT's Sequencer connection flow. The user can use their existing Google or email login, and Continue with ChatGPT when enabled. Never request a password, token, or verification code in chat.
3. Offer `open_sequencer_workspace` for projects, recent creations, and shot previews. Tools remain usable when the client has no sidebar support.
4. Ask for a brief or source image only if it is missing. Use available task-specific input without asking the user to repeat it.
5. Inspect the supported model inputs, request an upfront `quote_generation`, show the concrete plan and USD cost, and wait for approval before generating. Keep the selected model and approved budget through retries.
6. Generation uses the connected Sequencer account's existing allowance or balance. If unavailable, explain the entitlement requirement. You may link to `https://sequencer.media/plugin/account`, which is informational. Do not display subscription plans, promote upgrades, initiate checkout, or direct users to a purchase page. The user can still plan, get quotes, and work with existing media.
7. After a generation starts, use View Media to show progress and the completed result. Creations are already saved in the user's workspace. Offer refinement or the actual project link as the next step.
Keep instructions brief. Do not create a paid job to test installation. Do not collect the full conversation or expose private routing information.
Referenced files: 1 — file references are metadata, not package contents.
short-explainer2.25 KB
View saved version →
---
name: short-explainer
description: "Create and edit narrated explainer videos with an on-screen avatar in Sequencer, using consistent voice and appearance references, planned camera coverage, clip-by-clip production, and full-sequence continuity review. Use for presenter-led educational or explanatory videos, not live interactive avatars or faceless narration alone."
---
# Avatar Explainer Videos
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `short-explainer`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Topic, source material or script. Presenter and voice references or direction.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
spark-microdramas2.01 KB
View saved version →
---
name: spark-microdramas
description: "Turn your story into a connected series of vertical drama episodes. Run this workflow with Sequencer."
---
# Microdramas
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `spark-microdramas`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Story text or a premise, spoken language, and the requested episode scope.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
sphere-world2.02 KB
View saved version →
---
name: sphere-world
description: "Build an immersive spherical world with portals, custom rooms and a mobile fallback. Run this workflow with Sequencer."
---
# 360 Portal World
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `sphere-world`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: A hosted-world brief and real experiences or imagery for its portals.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
static-ads2.07 KB
View saved version →
---
name: static-ads
description: "Create static advertising images, banner ads or paid-social creative variations with exact copy and placement-specific layouts. Run this workflow with Sequencer."
---
# Create static ads
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `static-ads`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Product or brand references, verified offer and copy, placements and requested variants.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
ugc-ad-producer2.08 KB
View saved version →
---
name: ugc-ad-producer
description: "Create a new creator-led product ad, unboxing, demonstration or tutorial using consistent presenter and product references. Run this workflow with Sequencer."
---
# UGC Ad Producer
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `ugc-ad-producer`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Verified product references and the ad brief. Presenter reference or direction when visible.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
url-to-video1.97 KB
View saved version →
---
name: url-to-video
description: "Turn a public webpage into a video grounded in its verified copy and imagery. Run this workflow with Sequencer."
---
# URL to Video
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `url-to-video`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Exactly one usable public webpage URL.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
video-analysis2 KB
View saved version →
---
name: video-analysis
description: "Analyze a video or Sequencer edit, diagnose awkward animation and editing, and recommend specific timestamped improvements."
---
# Video analysis
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `video-analysis`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: An existing edit or accessible video to review.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
video-explainer2.13 KB
View saved version →
---
name: video-explainer
description: "Turn a topic, document or supplied explanation into a concise visual explainer in Sequencer with narration, diagrams and measured timing. Use for teaching a mechanism or idea; use existing News or Faceless skills for those specific formats."
---
# Video Explainer
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `video-explainer`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Topic, source material or an explanation to visualize.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
video-localization2.13 KB
View saved version →
---
name: video-localization
description: "Localize an existing video in Sequencer with translated captions, graphics and optional speech while preserving meaning and edit structure. Use for language adaptation, not voice changes unrelated to translation."
---
# Video Localization
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `video-localization`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Source video/edit and script or transcript, target locales and requested treatment.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
website-builder2.12 KB
View saved version →
---
name: website-builder
description: "Plan and build a responsive website or landing page using Sequencer-generated media and the site/file tools available to Spark. Use for a website deliverable; use Create Worldstream for interactive streaming experiences."
---
# Website Builder
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `website-builder`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Website purpose, content and requested pages or an existing site.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
workflow-building2.18 KB
View saved version →
---
name: workflow-building
description: "Build, connect, configure, and explain Sequencer node workflows. Use for the workflow canvas and reusable media pipelines, including edits to the open graph. Standalone media generation and film timelines use their own capabilities."
---
# Workflow Building
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `workflow-building`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: A workflow goal or requested graph change. An accessible workspace and an existing workflow target when editing.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.
write-ai-video-prompts1.22 KB
View saved version →
---
name: write-ai-video-prompts
description: "Turn an idea into clear image and motion prompts with camera, timing, and continuity notes. Run this workflow with Sequencer."
---
# Write AI Video Prompts
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `write-ai-video-prompts`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Expected inputs: A scene idea, references or shot brief to translate into prompts.
Return the requested writing or audit in the conversation. Workspace setup, generation and paid-production approval are required only for separately requested production work.
Referenced files: 1 — file references are metadata, not package contents.
write-better-stories1.21 KB
View saved version →
---
name: write-better-stories
description: "Write and refine dramatic scenes with clear goals, consequential choices, and meaningful reversals. Run this workflow with Sequencer."
---
# Write Better Stories
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `write-better-stories`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Expected inputs: A story brief, dramatic scene or sequence to write or audit.
Return the requested writing or audit in the conversation. Workspace setup, generation and paid-production approval are required only for separately requested production work.
Referenced files: 1 — file references are metadata, not package contents.
youtube-thumbnail2.05 KB
View saved version →
---
name: youtube-thumbnail
description: "Create YouTube thumbnails or vertical video covers from the actual video topic and supplied identity references. Run this workflow with Sequencer."
---
# YouTube Thumbnail
Run this workflow through the Sequencer MCP and its connected account. Use Sequencer tools for model discovery, quotes, generation and delivery. If the required tools or account connection are unavailable, follow the Sequencer setup flow at https://sequencer.media/plugin?source=chatgpt, then resume the request. Use another provider only when the user explicitly requests it.
An explanation or planning request authorizes only that scope.
Before planning, call `list_sequencer_skills`. This entrypoint describes `youtube-thumbnail`. Load that exact ID with `get_sequencer_skill` when it matches the user goal and is available. Follow the current returned instructions. If it is unavailable, explain the limitation and choose another workflow only when it fits the request.
Show completed videos and revisions with render_media. For an exported project, poll get_export to completed and use its exact renderMediaArgs. For standalone media use its returned mediaId. Keep download and project links as fallbacks. Avoid Markdown or HTML video embeds.
Expected inputs: Video topic, title or source video and any identity references.
When a proposal is required, keep the workflow’s own proposal and approval tools. For film/edit proposals save the full details with a short proposalSummary. Show the summary, a View full proposal link using reviewUrl, and ask for approval or changes here in chat. Use a host-provided approval-capable question tool when available; otherwise accept a chat reply. For film/edit proposals call set_proposal_status with the exact proposalId after explicit approval. Show full proposal text only when requested.
Use `get_model_catalog` for current models and `quote_generation` for concrete settings and cost. Obtain approval of the production scope and budget before paid work. Respect existing approvals. Report actual outputs and tool failures.
Referenced files: 1 — file references are metadata, not package contents.