SlideMaster
io.github.headdao/slidemaster-mcp
32 toolsauthSTDIOregistry active
Summary
This connects Claude to SlideMaster AI's course design platform, letting you generate complete presentation videos through conversational prompts. You give it a topic, duration, and audience, and it returns structured lesson plans with slides, narration scripts, TTS audio, and MP4 exports. The underlying service builds what they call "experience blueprints" that map out teaching rhythm and interaction points, then renders everything into a finished video with AI voiceover and subtitles. Reach for this when you need to go from subject matter to a polished instructional video without touching presentation or video editing software. It's built for corporate training, online courses, and lecture recording workflows where speed matters more than pixel perfect customization.
Keep your Mac awake while Claude Code and 40+ AI agents run. Sleeps when they're idle.
One time payment $9 →
Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →
Tools
Public tool metadata for what this MCP can expose to an agent.
32 tools
generate_outlineUse this when the user wants to plan a presentation. Generates slide titles and descriptions from a topic. Automatically creates a project and returns project_id — no need to call create_project separately.4 params
Use this when the user wants to plan a presentation. Generates slide titles and descriptions from a topic. Automatically creates a project and returns project_id — no need to call create_project separately.
Parameters* required
stylestringPresentation styleone of
professional · playful · minimalisttopicstringThe topic or prompt, e.g. 'Introduction to Machine Learning'
languagestringOutput language (default: 'zh-TW')one of
zh-TW · zh-CN · en · jaslides_countintegerNumber of slides (default: 8)
create_projectUse this to create a new empty project. Returns the project_id needed for all subsequent operations.3 params
Use this to create a new empty project. Returns the project_id needed for all subsequent operations.
Parameters* required
typestring'traditional' (slides) or 'flow' (screen recording). Default: 'traditional'one of
traditional · flowtitlestringProject title, e.g. 'Q3 Sales Report'
descriptionstringOptional project description
upload_initUse this to upload an existing PPTX or PDF file. Returns a pre-signed upload URL and project_id.2 params
Use this to upload an existing PPTX or PDF file. Returns a pre-signed upload URL and project_id.
Parameters* required
titlestringTitle for the uploaded presentation
filenamestringOriginal filename with extension, e.g. 'deck.pptx'
upload_completeUse this after the file has been uploaded to the pre-signed URL. Triggers server-side processing.1 params
Use this after the file has been uploaded to the pre-signed URL. Triggers server-side processing.
Parameters* required
project_idstringProject ID returned from upload_init
analyze_style_imageUpload a style reference image to analyze its visual style (colors, fonts, layout). Returns a style_guide object that can be passed to render_slides to apply this style to all slides. IMPORTANT: image_url must be a publicly accessible URL (e.g. hosted on Imgur, Cloudinary, or...1 params
Upload a style reference image to analyze its visual style (colors, fonts, layout). Returns a style_guide object that can be passed to render_slides to apply this style to all slides. IMPORTANT: image_url must be a publicly accessible URL (e.g. hosted on Imgur, Cloudinary, or...
Parameters* required
image_urlstringPublic URL of the style reference image (PNG, JPG, or any image URL)
render_slidesUse this to generate AI slide images for a project. Each slide gets a background image from the AI model. Optionally pass a style_guide (from analyze_style_image) to apply a consistent visual style. Each slide needs: title (displayed heading), description (visual scene descrip...5 params
Use this to generate AI slide images for a project. Each slide gets a background image from the AI model. Optionally pass a style_guide (from analyze_style_image) to apply a consistent visual style. Each slide needs: title (displayed heading), description (visual scene descrip...
Parameters* required
slidesarrayArray of {title, description} objects
dry_runbooleanIf true, only return estimated cost without actually rendering. Use to preview cost before committing.
image_sizestringResolution: '1K' | '2K' | '4K' (default: '2K')one of
1K · 2K · 4Kproject_idstringTarget project ID
style_guideobjectStyle guide object from analyze_style_image. Pass this to apply a consistent visual style to all slides.
generate_scriptUse this to generate a narration script for a single slide using AI.3 params
Use this to generate a narration script for a single slide using AI.
Parameters* required
slide_idstringSlide ID to generate a script for
regeneratebooleanForce regeneration if script exists (default: false)
prompt_stylestringNarration tone (default: 'professional')one of
professional · casual · educationalbatch_generate_scriptsUse this to generate narration scripts for all slides in a project in one batch.4 params
Use this to generate narration scripts for all slides in a project in one batch.
Parameters* required
slide_idsarrayOptional subset of slide IDs; omit for all slides
project_idstringProject ID to generate scripts for
regeneratebooleanForce regeneration if scripts exist (default: false)
prompt_stylestringNarration tone (default: 'professional')one of
professional · casual · educationalgenerate_ttsUse this to generate text-to-speech audio. Pass project_id for all slides or slide_id for one slide. Provider comparison: - **gemini**: Best quality for zh-TW/en/ja, natural intonation, recommended default. Voices: Zephyr, Puck, Charon, Kore, etc. - **cosyvoice**: Supports voi...5 params
Use this to generate text-to-speech audio. Pass project_id for all slides or slide_id for one slide. Provider comparison: - **gemini**: Best quality for zh-TW/en/ja, natural intonation, recommended default. Voices: Zephyr, Puck, Charon, Kore, etc. - **cosyvoice**: Supports voi...
Parameters* required
slide_idstringSlide ID — generates TTS for one slide
project_idstringProject ID — generates TTS for all slides
voice_namestringVoice ID from the provider. Use list_voices to browse.
tts_providerstringTTS provider. Only 'gemini', 'cosyvoice', or 'azure' are supported.one of
gemini · cosyvoice · azurespeaking_ratenumberSpeed multiplier, e.g. 0.8 slower, 1.2 faster (default: 1.0)
generate_videoUse this as the final step to compile slides + audio into an MP4 video. Poll check_status afterward.4 params
Use this as the final step to compile slides + audio into an MP4 video. Poll check_status afterward.
Parameters* required
fpsintegerFrames per second (default: 30)
forcebooleanForce regeneration if video exists (default: false)
project_idstringProject ID to generate video for
resolutionarray[width, height], e.g. [1920, 1080] (default: [1920, 1080])
list_projectsUse this when the user wants to see their existing projects. Returns a paginated project list.1 params
Use this when the user wants to see their existing projects. Returns a paginated project list.
Parameters* required
limitintegerMax results to return (default: 50, max: 100)
get_projectUse this to retrieve detailed metadata and status for a specific project.1 params
Use this to retrieve detailed metadata and status for a specific project.
Parameters* required
project_idstringProject ID to retrieve
update_projectUse this to update project settings like title, description, TTS provider, voice, or language.7 params
Use this to update project settings like title, description, TTS provider, voice, or language.
Parameters* required
titlestringNew project title
tts_voicestringVoice name/ID
project_idstringProject ID to update
descriptionstringNew project description
tts_languagestringLanguage code, e.g. 'zh-TW', 'en'
tts_providerstringTTS provider. Only 'gemini', 'cosyvoice', or 'azure' are supported.one of
gemini · cosyvoice · azuretts_speaking_ratenumberSpeed multiplier (default: 1.0)
delete_projectUse this to permanently delete a project and all its slides, audio, and video. Do not use for archiving.1 params
Use this to permanently delete a project and all its slides, audio, and video. Do not use for archiving.
Parameters* required
project_idstringProject ID to delete
list_slidesUse this to list all slides in a project with their titles, scripts, and render status.1 params
Use this to list all slides in a project with their titles, scripts, and render status.
Parameters* required
project_idstringProject ID to list slides for
update_slideUse this to edit a slide's title or narration script text.3 params
Use this to edit a slide's title or narration script text.
Parameters* required
titlestringNew slide title
scriptstringNew narration script text
slide_idstringSlide ID to update
delete_slideUse this to permanently remove a single slide from a project.1 params
Use this to permanently remove a single slide from a project.
Parameters* required
slide_idstringSlide ID to delete
check_statusUse this to poll processing progress after triggering render, TTS, or video generation. Status values: - "processing": Slides are being rendered (poll every 5-10 seconds) - "parsed": Slides are rendered and ready. Check scripts/audio progress in the response. - "generating_vid...1 params
Use this to poll processing progress after triggering render, TTS, or video generation. Status values: - "processing": Slides are being rendered (poll every 5-10 seconds) - "parsed": Slides are rendered and ready. Check scripts/audio progress in the response. - "generating_vid...
Parameters* required
project_idstringProject ID to check status for
export_projectUse this to get download URLs for all generated assets: slides, audio, and video files.1 params
Use this to get download URLs for all generated assets: slides, audio, and video files.
Parameters* required
project_idstringProject ID to export
export_pdfExport all rendered slides as a single PDF file (16:9 landscape, one slide per page). Returns { pdf_url, page_count, title, summary }. IMPORTANT: Give the user the pdf_url value — it is a signed download link that expires. Do NOT construct GCS URLs manually.1 params
Export all rendered slides as a single PDF file (16:9 landscape, one slide per page). Returns { pdf_url, page_count, title, summary }. IMPORTANT: Give the user the pdf_url value — it is a signed download link that expires. Do NOT construct GCS URLs manually.
Parameters* required
project_idstringProject ID to export as PDF
export_pptxExport all rendered slides as a single PPTX (PowerPoint) file. Returns { pptx_url, page_count, title, summary }. IMPORTANT: Give the user the pptx_url value — it is a signed download link that expires.1 params
Export all rendered slides as a single PPTX (PowerPoint) file. Returns { pptx_url, page_count, title, summary }. IMPORTANT: Give the user the pptx_url value — it is a signed download link that expires.
Parameters* required
project_idstringProject ID to export as PPTX
export_evercamExport project as an EverCam offline HTML5 playable package (ZIP). Requires video to be generated first. This is async — call evercam_status to poll progress.1 params
Export project as an EverCam offline HTML5 playable package (ZIP). Requires video to be generated first. This is async — call evercam_status to poll progress.
Parameters* required
project_idstringProject ID to export as EverCam
evercam_statusPoll EverCam export progress. Returns status (processing/completed/failed/not_started) and download_url when completed.1 params
Poll EverCam export progress. Returns status (processing/completed/failed/not_started) and download_url when completed.
Parameters* required
project_idstringProject ID to check EverCam export status
list_voicesBrowse available TTS voices. Use this before generate_tts to help the user pick a voice. Providers: gemini (best quality, zh-TW/en/ja), cosyvoice (voice cloning, zh-TW/zh-CN), azure (40+ languages, widest coverage).1 params
Browse available TTS voices. Use this before generate_tts to help the user pick a voice. Providers: gemini (best quality, zh-TW/en/ja), cosyvoice (voice cloning, zh-TW/zh-CN), azure (40+ languages, widest coverage).
Parameters* required
providerstringFilter by provider. Omit for all providers.one of
gemini · cosyvoice · azureget_credit_balanceCheck remaining API credits before starting expensive operations. Call this before large projects (>10 slides) to avoid mid-job failures. Quick reference: 2,000 credits ≈ 14 slides at 2K resolution (slides only, no TTS/video). Per-slide costs: 1K=42, 2K=140, 4K=250 credits. TT...
Check remaining API credits before starting expensive operations. Call this before large projects (>10 slides) to avoid mid-job failures. Quick reference: 2,000 credits ≈ 14 slides at 2K resolution (slides only, no TTS/video). Per-slide costs: 1K=42, 2K=140, 4K=250 credits. TT...
No parameter schema in public metadata yet.
get_credit_transactionsView recent credit transaction history (purchases, consumption, monthly resets, refunds). Use this when the user asks about their billing history or wants to know what consumed their credits.1 params
View recent credit transaction history (purchases, consumption, monthly resets, refunds). Use this when the user asks about their billing history or wants to know what consumed their credits.
Parameters* required
limitintegerNumber of transactions to return (default: 20, max: 100)
estimate_costEstimate credit cost BEFORE running expensive operations. Returns per-operation breakdown, total, current balance, and whether you have enough. Always call this before render_slides or topic_to_video on large projects. YouTube cost is dynamic: min 30 credits, actual = max(30,...3 params
Estimate credit cost BEFORE running expensive operations. Returns per-operation breakdown, total, current balance, and whether you have enough. Always call this before render_slides or topic_to_video on large projects. YouTube cost is dynamic: min 30 credits, actual = max(30,...
Parameters* required
image_sizestringResolution (default: 2K)one of
1K · 2K · 4KoperationsarrayOperations to estimate. 'full_pipeline' = render + tts + video. 'youtube' = YouTube-to-outline (min 30 credits). Default: full_pipeline.
slides_countintegerNumber of slides (default: 8)
get_project_contextGet a complete snapshot of a project in one call: metadata, all slides with script/audio status, progress summary, and what to do next. Call this after any long-running operation to orient yourself before deciding the next action.1 params
Get a complete snapshot of a project in one call: metadata, all slides with script/audio status, progress summary, and what to do next. Call this after any long-running operation to orient yourself before deciding the next action.
Parameters* required
project_idstringProject ID to inspect
topic_to_videoUse this for one-click presentation creation from a topic. Default: slides only. Set include_scripts=true for narration, include_video=true for full MP4. Poll check_status afterward.10 params
Use this for one-click presentation creation from a topic. Default: slides only. Set include_scripts=true for narration, include_video=true for full MP4. Poll check_status afterward.
Parameters* required
stylestringPresentation styleone of
professional · playful · minimalisttopicstringThe topic, e.g. 'Introduction to Machine Learning'
dry_runbooleanIf true, only return estimated cost without executing. Use to preview total pipeline cost.
languagestringOutput language (default: 'zh-TW')one of
zh-TW · zh-CN · en · jatts_voicestringVoice name/ID from the chosen provider
image_sizestringResolution (default: '2K')one of
1K · 2K · 4Kslides_countintegerNumber of slides (default: 8)
tts_providerstringTTS providerone of
gemini · cosyvoice · azureinclude_videobooleanGenerate TTS + compile MP4, implies include_scripts (default: false)
include_scriptsbooleanGenerate narration scripts (default: false)
get_preferencesGet saved user preferences (style, voice, language, etc). Call this at the START of every conversation to auto-apply the user's past choices. Returns null fields if no preferences saved yet.
Get saved user preferences (style, voice, language, etc). Call this at the START of every conversation to auto-apply the user's past choices. Returns null fields if no preferences saved yet.
No parameter schema in public metadata yet.
save_preferencesSave user preferences for future sessions. Call this when the user explicitly states a preference (e.g. 'I prefer playful style', 'always use Azure voice'). Only send fields that should be updated — existing fields are preserved.6 params
Save user preferences for future sessions. Call this when the user explicitly states a preference (e.g. 'I prefer playful style', 'always use Azure voice'). Only send fields that should be updated — existing fields are preserved.
Parameters* required
stylestringPresentation styleone of
professional · playful · minimalistlanguagestringPreferred languageone of
zh-TW · zh-CN · en · jatts_voicestringPreferred voice name
image_sizestringPreferred image resolutionone of
1K · 2K · 4Kslides_countintegerDefault number of slides
tts_providerstringPreferred TTS providerone of
gemini · cosyvoice · azureyoutube_to_outlineAnalyze a YouTube video and generate a structured course outline. Extracts the transcript (or uses video analysis as fallback) and reorganizes it into a teaching outline. Cost: min 30 credits, actual = max(30, ceil(video_duration_minutes) x 2). A 5-min video costs 30 credits,...2 params
Analyze a YouTube video and generate a structured course outline. Extracts the transcript (or uses video analysis as fallback) and reorganizes it into a teaching outline. Cost: min 30 credits, actual = max(30, ceil(video_duration_minutes) x 2). A 5-min video costs 30 credits,...
Parameters* required
urlstringYouTube video URL
tonestringOutput tone: 專業 (professional), 活潑 (lively), 幽默 (humorous), 嚴謹 (rigorous). Default: 專業one of
專業 · 活潑 · 幽默 · 嚴謹Featured
Keep your Mac awake while Claude Code and 40+ AI agents run. Sleeps when they're idle.
One time payment $9 →Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →Configuration
SLIDEMASTER_API_KEY*secretYour SlideMaster API key (get one at https://slidemaster.tw/settings/api-keys)
SLIDEMASTER_API_BASEOverride API base URL (default: https://api.slidemaster.tw/api/v1/public)
Categories
Registryactive
Package@slidemaster/mcp-server
TransportSTDIO
AuthRequired
UpdatedFeb 28, 2026
