Draftless is an MCP server published by dev.draftless, hosted at mcp.draftless.dev. When last checked on 2026-11-21 it was working: it answers and lists its tools. It declares 24 tools, including discover, plan, generate, ads, characters.
AI creative studio: generate image, video, voice and music, finish ads, publish to social.
5 connections through 5 shared values. Tap any node.
Discover tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_capabilities: List AI node types with descriptions and credit costs, plus the available models per modality (image, video, audio, text, 3d) with pricing a - list_animation_actions: Browse the 678 animation presets a rigged mesh can be animated with. - recommend_models: Recommend the best AI models for a creative goal, ranked over the live model catalog (tags, name, description) — like asking 'which model sh - recommend_pipeline: Get best-practice recommendations for a creative task. - list_video_edit_operations: Report which `apply_video_edit` operations are actually enabled on THIS deployment, and which are switched off with the flag that gates each - get_workflow_instructions: Discover and load Draftless's guided multi-step workflows. - preflight_check: Confirm connectivity, authentication, and credit balance before spending on a generation. - validate_params: Check that a node_type + inputs pair is well-formed before calling execute_node. - describe_action: Returns the exact input JSON schema, one working example and the credit cost for any action (e.g.
| Parameter | Type | Description |
|---|---|---|
| input |
None yet. Changes appear after a server's second daily check.
Squares are this server's values; dots are other servers sharing them. A shared address or publisher is a lead, not proof of a common operator.
Inferred from the host name and headers: what a server runs on, not who runs it.
| object |
| Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "discover",
"title": "Discover",
"description": "Discover tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_capabilities: List AI node types with descriptions and credit costs, plus the available models per modality (image, video, audio, text, 3d) with pricing a\n- list_animation_actions: Browse the 678 animation presets a rigged mesh can be animated with.\n- recommend_models: Recommend the best AI models for a creative goal, ranked over the live model catalog (tags, name, description) — like asking 'which model sh\n- recommend_pipeline: Get best-practice recommendations for a creative task.\n- list_video_edit_operations: Report which `apply_video_edit` operations are actually enabled on THIS deployment, and which are switched off with the flag that gates each\n- get_workflow_instructions: Discover and load Draftless's guided multi-step workflows.\n- preflight_check: Confirm connectivity, authentication, and credit balance before spending on a generation.\n- validate_params: Check that a node_type + inputs pair is well-formed before calling execute_node.\n- describe_action: Returns the exact input JSON schema, one working example and the credit cost for any action (e.g.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_capabilities",
"list_animation_actions",
"recommend_models",
"recommend_pipeline",
"list_video_edit_operations",
"get_workflow_instructions",
"preflight_check",
"validate_params",
"describe_action"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": true,
"destructiveHint": false
}
}Plan tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - plan_creative: Plan any creative deliverable — an ad, UGC reel, storyboard, cinematic shot list (Director mode), lipsync sequence, product photoshoot, conc - execute_plan: Materialize a storyboard plan (as returned by plan_creative (goal:'storyboard')) into real assets: generate one hero image per beat and subm - apply_prompt_enhancement: Rewrite a short video idea into a model-ready shot prompt, tuned for the target model, aspect ratio, duration and platform. - score_virality: Score a short-form concept for virality (0-100) using a 6-axis rubric (hook, pattern-interrupt, emotion, clarity, payoff, platform-fit). - plan_faceless_series: Plan a 'What if…' style series of 1-4 narrated faceless episodes from one premise: an arc, a cliffhanger at the end of every episode but the - generate_series_script: PAID, two-phase.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "plan",
"title": "Plan",
"description": "Plan tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- plan_creative: Plan any creative deliverable — an ad, UGC reel, storyboard, cinematic shot list (Director mode), lipsync sequence, product photoshoot, conc\n- execute_plan: Materialize a storyboard plan (as returned by plan_creative (goal:'storyboard')) into real assets: generate one hero image per beat and subm\n- apply_prompt_enhancement: Rewrite a short video idea into a model-ready shot prompt, tuned for the target model, aspect ratio, duration and platform.\n- score_virality: Score a short-form concept for virality (0-100) using a 6-axis rubric (hook, pattern-interrupt, emotion, clarity, payoff, platform-fit).\n- plan_faceless_series: Plan a 'What if…' style series of 1-4 narrated faceless episodes from one premise: an arc, a cliffhanger at the end of every episode but the\n- generate_series_script: PAID, two-phase.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"plan_creative",
"execute_plan",
"apply_prompt_enhancement",
"score_virality",
"plan_faceless_series",
"generate_series_script"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Generate tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - generate_one_take_ad: Proven one-take ad method, step by step (ads.one_take is the board-gated flow). - generate_image: Generate an image from a text prompt as the signed-in user. - generate_video: Submit a video generation job as the signed-in user. - generate_audio: Generate audio as the signed-in user and save it to the Asset Library. - execute_node: Execute a single AI node as the signed-in user — generate an image, video, audio, 3D model, SVG, or text; analyze/relight/outpaint/inpaint/u - execute_chain: Run up to 10 AI nodes in sequence, wiring each step's outputs into the next. - generate_batch: Run up to 10 independent AI node executions concurrently as the signed-in user. - generate_image_layers: Separate a single image into individually editable layers (background plate plus each distinct object, each a transparent PNG with its own b - generate_layered_poster: NOT AVAILABLE YET: switched off until the /layers poster editor ships (it answers 'not available yet'). - generate_3d: Submit an asynchronous 3D job as the signed-in user: image-to-3D, text-to-3D, multi-view-to-3D, rigging of an existing mesh, or text-to-moti - generate_headshots: Generate one professional AI headshot of a person and save it to the user's Asset Library. - generate_try_on: Virtually try a garment on a person image and save the result to the user's Asset Library. - plan_saas_video: Writes a SaaS launch film for a product (the same engine as Draftless SaaS Video): the copy, the product UI it shows, its motion, and the EX - create_saas_video: Renders a film planned by plan_saas_video. - get_saas_video: Status of a create_saas_video render: { status, progress, resultUrl (signed MP4 when done), error, refunded }. - plan_story_film: One prompt → a story-led product launch film: two named people in two places, the product's REAL captured pages and components driven by a n - get_story_plan: The plan started by plan_story_film: { status: 'planning' } while the story is written, then { status: 'ready', ok, storyboard, credits, cei - create_story_film: Starts the film planned by plan_story_film: pass its plan_id (once get_story_plan says ready), accepted_credits = the plan's `credits` and a - get_story_film: Status of a create_story_film run: { status, items, director: { phase, error, log, filmUrl } }. - create_story_rerender: For a create_story_film run whose edit or render FAILED after its takes were made: cuts the stored takes again and renders, with no new gene
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "generate",
"title": "Generate",
"description": "Generate tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- generate_one_take_ad: Proven one-take ad method, step by step (ads.one_take is the board-gated flow).\n- generate_image: Generate an image from a text prompt as the signed-in user.\n- generate_video: Submit a video generation job as the signed-in user.\n- generate_audio: Generate audio as the signed-in user and save it to the Asset Library.\n- execute_node: Execute a single AI node as the signed-in user — generate an image, video, audio, 3D model, SVG, or text; analyze/relight/outpaint/inpaint/u\n- execute_chain: Run up to 10 AI nodes in sequence, wiring each step's outputs into the next.\n- generate_batch: Run up to 10 independent AI node executions concurrently as the signed-in user.\n- generate_image_layers: Separate a single image into individually editable layers (background plate plus each distinct object, each a transparent PNG with its own b\n- generate_layered_poster: NOT AVAILABLE YET: switched off until the /layers poster editor ships (it answers 'not available yet').\n- generate_3d: Submit an asynchronous 3D job as the signed-in user: image-to-3D, text-to-3D, multi-view-to-3D, rigging of an existing mesh, or text-to-moti\n- generate_headshots: Generate one professional AI headshot of a person and save it to the user's Asset Library.\n- generate_try_on: Virtually try a garment on a person image and save the result to the user's Asset Library.\n- plan_saas_video: Writes a SaaS launch film for a product (the same engine as Draftless SaaS Video): the copy, the product UI it shows, its motion, and the EX\n- create_saas_video: Renders a film planned by plan_saas_video.\n- get_saas_video: Status of a create_saas_video render: { status, progress, resultUrl (signed MP4 when done), error, refunded }.\n- plan_story_film: One prompt → a story-led product launch film: two named people in two places, the product's REAL captured pages and components driven by a n\n- get_story_plan: The plan started by plan_story_film: { status: 'planning' } while the story is written, then { status: 'ready', ok, storyboard, credits, cei\n- create_story_film: Starts the film planned by plan_story_film: pass its plan_id (once get_story_plan says ready), accepted_credits = the plan's `credits` and a\n- get_story_film: Status of a create_story_film run: { status, items, director: { phase, error, log, filmUrl } }.\n- create_story_rerender: For a create_story_film run whose edit or render FAILED after its takes were made: cuts the stored takes again and renders, with no new gene",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"generate_one_take_ad",
"generate_image",
"generate_video",
"generate_audio",
"execute_node",
"execute_chain",
"generate_batch",
"generate_image_layers",
"generate_layered_poster",
"generate_3d",
"generate_headshots",
"generate_try_on",
"plan_saas_video",
"create_saas_video",
"get_saas_video",
"plan_story_film",
"get_story_plan",
"create_story_film",
"get_story_film",
"create_story_rerender"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Ads tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - one_take: Board-approved ad flow.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "ads",
"title": "Ads",
"description": "Ads tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- one_take: Board-approved ad flow.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"one_take"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Characters tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - influencer_options: List the AI Influencer builder's trait groups and options, looks, sheet layouts, price tiers (draft|pro) and limits. - influencer_presets: List the starter AI Influencer recipes (name, traits, look, prompt) with a recreateUrl that opens the builder prefilled. - create_influencer_sheet: Design an original adult AI Influencer from traits and generate its character sheet (1-4 sheets). - save_sheet_as_element: Turn a finished AI Influencer sheet (a `job_item_id` from characters.get_identity_job) into a People element: crops the portrait and full-bo - run_identity_shoot: Photo set of ONE ready People element (same face across scenes): packs photodump (8), city-week (8), studio (4), headshots (6) or custom `lo - get_identity_job: Poll every 10-20 seconds until status is done, partial or failed. - list_characters: List the signed-in user's Draftless characters (Soul-ID identities) used for consistent talent across UGC, ads, and storyboards. - create_character: Create a new Draftless character (Soul-ID identity) from pre-hosted reference image URLs. - update_character: Update an existing Draftless character (Soul-ID identity). - delete_character: Permanently delete a Draftless character (Soul-ID identity). - render_avatar: Character-locked talking avatar: takes a character_id and either audio_url or script_text. - render_lipsync: Lipsync a source VIDEO to an audio track.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "characters",
"title": "Characters",
"description": "Characters tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- influencer_options: List the AI Influencer builder's trait groups and options, looks, sheet layouts, price tiers (draft|pro) and limits.\n- influencer_presets: List the starter AI Influencer recipes (name, traits, look, prompt) with a recreateUrl that opens the builder prefilled.\n- create_influencer_sheet: Design an original adult AI Influencer from traits and generate its character sheet (1-4 sheets).\n- save_sheet_as_element: Turn a finished AI Influencer sheet (a `job_item_id` from characters.get_identity_job) into a People element: crops the portrait and full-bo\n- run_identity_shoot: Photo set of ONE ready People element (same face across scenes): packs photodump (8), city-week (8), studio (4), headshots (6) or custom `lo\n- get_identity_job: Poll every 10-20 seconds until status is done, partial or failed.\n- list_characters: List the signed-in user's Draftless characters (Soul-ID identities) used for consistent talent across UGC, ads, and storyboards.\n- create_character: Create a new Draftless character (Soul-ID identity) from pre-hosted reference image URLs.\n- update_character: Update an existing Draftless character (Soul-ID identity).\n- delete_character: Permanently delete a Draftless character (Soul-ID identity).\n- render_avatar: Character-locked talking avatar: takes a character_id and either audio_url or script_text.\n- render_lipsync: Lipsync a source VIDEO to an audio track.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"influencer_options",
"influencer_presets",
"create_influencer_sheet",
"save_sheet_as_element",
"run_identity_shoot",
"get_identity_job",
"list_characters",
"create_character",
"update_character",
"delete_character",
"render_avatar",
"render_lipsync"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": true
}
}Jobs tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - get_job: Look up a single asynchronous job (e.g. - list_jobs: List the signed-in user's asynchronous jobs (e.g. - get_video_status: Check the status of a video generation job submitted with `generate_video`. - get_jobs_status: Long-poll up to 12 async jobs together and return once every one has reached a terminal state (completed / failed / cancelled), or when the - get_3d_status: Check the status of a 3D mesh generation job submitted with `generate_3d`. - get_image_layers_status: Check the status of an image separation submitted with `generate_image_layers`. - get_music_status: Poll a music job returned by generate_audio (kind 'music') with status 'processing'.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "jobs",
"title": "Jobs",
"description": "Jobs tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- get_job: Look up a single asynchronous job (e.g.\n- list_jobs: List the signed-in user's asynchronous jobs (e.g.\n- get_video_status: Check the status of a video generation job submitted with `generate_video`.\n- get_jobs_status: Long-poll up to 12 async jobs together and return once every one has reached a terminal state (completed / failed / cancelled), or when the \n- get_3d_status: Check the status of a 3D mesh generation job submitted with `generate_3d`.\n- get_image_layers_status: Check the status of an image separation submitted with `generate_image_layers`.\n- get_music_status: Poll a music job returned by generate_audio (kind 'music') with status 'processing'.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"get_job",
"list_jobs",
"get_video_status",
"get_jobs_status",
"get_3d_status",
"get_image_layers_status",
"get_music_status"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Assets tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_assets: List the signed-in user's Asset Library entries (generated and uploaded images, videos, and audio) with their URLs. - import_media_url: Fetch a public media URL (image, video, or audio) and save it into the signed-in user's Asset Library. - upload_file: Upload local file bytes to Draftless storage in ONE call by passing them base64-encoded, and get back a signed URL (valid 24h) usable as `im - open_upload: Opens the upload view so the user can drag-drop or pick local files that go straight to their Draftless storage (then finalized into assets) - create_upload_url: Create a short-lived signed upload URL so an MCP client can PUT local file bytes directly to Draftless storage without going through a URL. - finalize_upload: After PUT-ing bytes to a signed upload URL from create_upload_url, call this with the same object_path to verify the file arrived and get a - create_upload_urls: Batch form of create_upload_url: sign up to 50 uploads in one call. - finalize_uploads: Batch form of finalize_upload: verify up to 50 uploaded objects and get a 24h signed read URL for each, in order ({ index, object_path, url - get_url: Mint a fresh 24h signed URL for media you own. - create_product_capture: Open a PUBLIC product URL in Draftless's headless browser, dismiss cookie banners, screenshot the landing page plus up to 3 distinct pages f - get_product_capture: Status of a create_product_capture job.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "assets",
"title": "Assets",
"description": "Assets tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_assets: List the signed-in user's Asset Library entries (generated and uploaded images, videos, and audio) with their URLs.\n- import_media_url: Fetch a public media URL (image, video, or audio) and save it into the signed-in user's Asset Library.\n- upload_file: Upload local file bytes to Draftless storage in ONE call by passing them base64-encoded, and get back a signed URL (valid 24h) usable as `im\n- open_upload: Opens the upload view so the user can drag-drop or pick local files that go straight to their Draftless storage (then finalized into assets)\n- create_upload_url: Create a short-lived signed upload URL so an MCP client can PUT local file bytes directly to Draftless storage without going through a URL.\n- finalize_upload: After PUT-ing bytes to a signed upload URL from create_upload_url, call this with the same object_path to verify the file arrived and get a \n- create_upload_urls: Batch form of create_upload_url: sign up to 50 uploads in one call.\n- finalize_uploads: Batch form of finalize_upload: verify up to 50 uploaded objects and get a 24h signed read URL for each, in order ({ index, object_path, url \n- get_url: Mint a fresh 24h signed URL for media you own.\n- create_product_capture: Open a PUBLIC product URL in Draftless's headless browser, dismiss cookie banners, screenshot the landing page plus up to 3 distinct pages f\n- get_product_capture: Status of a create_product_capture job.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_assets",
"import_media_url",
"upload_file",
"open_upload",
"create_upload_url",
"finalize_upload",
"create_upload_urls",
"finalize_uploads",
"get_url",
"create_product_capture",
"get_product_capture"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Image Edit tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - upscale_image: Upscale an image to higher resolution. - remove_background: Remove the background from an image and return a transparent PNG. - outpaint_image: Expand an image's canvas by a percentage on each side, filling the new area with AI-generated content. - reframe_image: Reframe an image to a new aspect ratio by outpainting new content on the sides (never cropping). - apply_relight: Relight an existing image: light direction, intensity, warmth and ambient fill, keeping subject and composition. - edit_image_inpaint: Edit an image inside a masked region: replace, remove, or restyle pixels covered by the mask while keeping the rest untouched. - composite_image: Add an image overlay (logo / badge / watermark), a text caption/label (true-type Inter, not bitmap), and/or a designed 'Powered by <Brand>' - add_badge: The result is saved to your asset library and returned as a durable URL.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "image_edit",
"title": "Image Edit",
"description": "Image Edit tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- upscale_image: Upscale an image to higher resolution.\n- remove_background: Remove the background from an image and return a transparent PNG.\n- outpaint_image: Expand an image's canvas by a percentage on each side, filling the new area with AI-generated content.\n- reframe_image: Reframe an image to a new aspect ratio by outpainting new content on the sides (never cropping).\n- apply_relight: Relight an existing image: light direction, intensity, warmth and ambient fill, keeping subject and composition.\n- edit_image_inpaint: Edit an image inside a masked region: replace, remove, or restyle pixels covered by the mask while keeping the rest untouched.\n- composite_image: Add an image overlay (logo / badge / watermark), a text caption/label (true-type Inter, not bitmap), and/or a designed 'Powered by <Brand>' \n- add_badge: The result is saved to your asset library and returned as a durable URL.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"upscale_image",
"remove_background",
"outpaint_image",
"reframe_image",
"apply_relight",
"edit_image_inpaint",
"composite_image",
"add_badge"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Video Edit tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - apply_video_edit: Video-to-video edit of an existing clip. - apply_reshoot: Doppelganger edits ONE existing video using reference images, as the signed-in user. - restyle_video: Redraw ONE video (3-30 s) in a named look (claymation, anime, pixel art, yarn, toy CGI, painterly 3D, gouache, comic…), keeping its motion, - multiply_ad: Ad Multiplier: ONE ad clip -> 2-12 variants, one per row; each row swaps product, person, location or outfit for its own 1-4 reference image - upscale_video: Submit an asynchronous video upscale job as the signed-in user. - motion_control: Motion work on an existing shot, selected by `mode`. - motion_recast: Plan or execute a camera-motion recast. - generate_frame_interpolation: Generate in-between STILL frames from two keyframes (one AI image edit per in-between frame). - analyze_video: Analyze a video (URL) with Gemini 2.5 Pro and return a scene-by-scene breakdown: shot type, camera motion, lighting, subjects, dialogue, and - clip_long_video: Turn a long video into N short-form clips. - dub_video: Translate a video or voiceover into another language in one step. - apply_speed_ramp: Speed-ramp an existing clip on our ffmpeg (no model): eased ramps, motion-interpolated slow motion, tearing-checked. - apply_deflicker: Remove AI brightness flicker on our ffmpeg: several settings are measured and the best one that keeps >= 95 % sharpness ships; never a clip
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "video_edit",
"title": "Video Edit",
"description": "Video Edit tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- apply_video_edit: Video-to-video edit of an existing clip.\n- apply_reshoot: Doppelganger edits ONE existing video using reference images, as the signed-in user.\n- restyle_video: Redraw ONE video (3-30 s) in a named look (claymation, anime, pixel art, yarn, toy CGI, painterly 3D, gouache, comic…), keeping its motion, \n- multiply_ad: Ad Multiplier: ONE ad clip -> 2-12 variants, one per row; each row swaps product, person, location or outfit for its own 1-4 reference image\n- upscale_video: Submit an asynchronous video upscale job as the signed-in user.\n- motion_control: Motion work on an existing shot, selected by `mode`.\n- motion_recast: Plan or execute a camera-motion recast.\n- generate_frame_interpolation: Generate in-between STILL frames from two keyframes (one AI image edit per in-between frame).\n- analyze_video: Analyze a video (URL) with Gemini 2.5 Pro and return a scene-by-scene breakdown: shot type, camera motion, lighting, subjects, dialogue, and\n- clip_long_video: Turn a long video into N short-form clips.\n- dub_video: Translate a video or voiceover into another language in one step.\n- apply_speed_ramp: Speed-ramp an existing clip on our ffmpeg (no model): eased ramps, motion-interpolated slow motion, tearing-checked.\n- apply_deflicker: Remove AI brightness flicker on our ffmpeg: several settings are measured and the best one that keeps >= 95 % sharpness ships; never a clip ",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"apply_video_edit",
"apply_reshoot",
"restyle_video",
"multiply_ad",
"upscale_video",
"motion_control",
"motion_recast",
"generate_frame_interpolation",
"analyze_video",
"clip_long_video",
"dub_video",
"apply_speed_ramp",
"apply_deflicker"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Review tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - create_video_review: Watch a rendered video and return a grounded quality verdict: a 0-100 score across six axes (hook, motion, fidelity, framing, on-brief, fini - get_video_review: Fetch the result of a `create_video_review` submission. - list_video_moments: Returns matching MOMENTS (time-slices) with a timecode ref, a relevance score, a text snippet and a thumbnail.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "review",
"title": "Review",
"description": "Review tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- create_video_review: Watch a rendered video and return a grounded quality verdict: a 0-100 score across six axes (hook, motion, fidelity, framing, on-brief, fini\n- get_video_review: Fetch the result of a `create_video_review` submission.\n- list_video_moments: Returns matching MOMENTS (time-slices) with a timecode ref, a relevance score, a text snippet and a thumbnail.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"create_video_review",
"get_video_review",
"list_video_moments"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Finish tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - composite_video: Assemble several video, image and audio clips into ONE finished MP4, rendered server-side. - get_composite_status: Check a composite render submitted with `composite_video`. - composite_text_on_video: Burn text onto an existing MP4 — captions, titles, watermarks, lower-thirds — and get back a durable file. - get_overlay_status: Check a text-overlay job submitted with `composite_text_on_video` a finish job submitted with `apply_finish`, or a speed-ramp / deflicker jo - apply_finish: Give an existing clip a finished, shot-on-a-camera feel and a proper delivery encode. - apply_music_bed: Lay a music track UNDER an existing MP4's own audio (voiceover, native speech) and get back the same video with mixed sound — what composite - transcribe_audio: Transcribe a public audio or video URL (≤25 MB, ≤ ~30 min) into text with timed segments (start/end seconds). - generate_captions: Transcribe a public audio or video URL and return timed captions in SRT, WebVTT, or JSON. - create_faceless_package: Write YouTube packaging for a narrated video: title options, description, tags, chapters snapped to shot boundaries, and thumbnail concepts.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "finish",
"title": "Finish",
"description": "Finish tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- composite_video: Assemble several video, image and audio clips into ONE finished MP4, rendered server-side.\n- get_composite_status: Check a composite render submitted with `composite_video`.\n- composite_text_on_video: Burn text onto an existing MP4 — captions, titles, watermarks, lower-thirds — and get back a durable file.\n- get_overlay_status: Check a text-overlay job submitted with `composite_text_on_video` a finish job submitted with `apply_finish`, or a speed-ramp / deflicker jo\n- apply_finish: Give an existing clip a finished, shot-on-a-camera feel and a proper delivery encode.\n- apply_music_bed: Lay a music track UNDER an existing MP4's own audio (voiceover, native speech) and get back the same video with mixed sound — what composite\n- transcribe_audio: Transcribe a public audio or video URL (≤25 MB, ≤ ~30 min) into text with timed segments (start/end seconds).\n- generate_captions: Transcribe a public audio or video URL and return timed captions in SRT, WebVTT, or JSON.\n- create_faceless_package: Write YouTube packaging for a narrated video: title options, description, tags, chapters snapped to shot boundaries, and thumbnail concepts.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"composite_video",
"get_composite_status",
"composite_text_on_video",
"get_overlay_status",
"apply_finish",
"apply_music_bed",
"transcribe_audio",
"generate_captions",
"create_faceless_package"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Voices tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - create_voice: Save a reusable voice identity from a consented audio sample, so the same voice can speak across generate_audio, render_lipsync, and render_ - list_voices: List the signed-in user's saved voice identities (plus any shared by their organization). - update_voice: Rename or re-tag a saved voice identity, or link it to a character. - delete_voice: Permanently delete a saved voice identity and release its provider-side clone (freeing the ElevenLabs voice slot for enrolled voices). - voice_change: Speech-to-speech: re-perform an existing recording in one of your saved voices, preserving the original timing, emphasis, and emotion (unlik
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "voices",
"title": "Voices",
"description": "Voices tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- create_voice: Save a reusable voice identity from a consented audio sample, so the same voice can speak across generate_audio, render_lipsync, and render_\n- list_voices: List the signed-in user's saved voice identities (plus any shared by their organization).\n- update_voice: Rename or re-tag a saved voice identity, or link it to a character.\n- delete_voice: Permanently delete a saved voice identity and release its provider-side clone (freeing the ElevenLabs voice slot for enrolled voices).\n- voice_change: Speech-to-speech: re-perform an existing recording in one of your saved voices, preserving the original timing, emphasis, and emotion (unlik",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"create_voice",
"list_voices",
"update_voice",
"delete_voice",
"voice_change"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": true
}
}Presets tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_style_presets: List Draftless style presets. - apply_style_preset: Compose a Draftless style preset with a subject/image into a ready-to-run payload. - list_reel_templates: List the curated Reels templates from the Draftless Reels studio: beat sheet, captions, aspect ratio and the video model each one runs on. - list_faceless_templates: List the Faceless Studio templates for narrated YouTube-style videos: tone, runtime, aspect ratio, visual mode (clip or still), music brief - list_faceless_styles: List the Faceless Studio art styles. - list_explainer_presets: List the 22 Faceless explainer presets. - list_poster_presets: List the 12 layered-poster looks. - list_product_shot_presets: Browse the 282 product-shot presets: scene recipes that put the user's OWN product photo into a styled shot (studio white, lifestyle, outdoo - resolve_explainer_preset: Resolve one Faceless explainer preset. - list_saas_video_templates: List deterministic, editable SaaS launch-video templates, including their semantic product-kit slots, scene timing, format support and audio - list_brand_kits: List the caller's brand kits (id, name, updated_at).
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "presets",
"title": "Presets",
"description": "Presets tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_style_presets: List Draftless style presets.\n- apply_style_preset: Compose a Draftless style preset with a subject/image into a ready-to-run payload.\n- list_reel_templates: List the curated Reels templates from the Draftless Reels studio: beat sheet, captions, aspect ratio and the video model each one runs on.\n- list_faceless_templates: List the Faceless Studio templates for narrated YouTube-style videos: tone, runtime, aspect ratio, visual mode (clip or still), music brief \n- list_faceless_styles: List the Faceless Studio art styles.\n- list_explainer_presets: List the 22 Faceless explainer presets.\n- list_poster_presets: List the 12 layered-poster looks.\n- list_product_shot_presets: Browse the 282 product-shot presets: scene recipes that put the user's OWN product photo into a styled shot (studio white, lifestyle, outdoo\n- resolve_explainer_preset: Resolve one Faceless explainer preset.\n- list_saas_video_templates: List deterministic, editable SaaS launch-video templates, including their semantic product-kit slots, scene timing, format support and audio\n- list_brand_kits: List the caller's brand kits (id, name, updated_at).",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_style_presets",
"apply_style_preset",
"list_reel_templates",
"list_faceless_templates",
"list_faceless_styles",
"list_explainer_presets",
"list_poster_presets",
"list_product_shot_presets",
"resolve_explainer_preset",
"list_saas_video_templates",
"list_brand_kits"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Playbooks tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_playbooks: List Draftless playbooks — named, ready-made workflows (Ad Strategist, Hook Multiplier, Effect Recreate…) grouped into bundles. - get_playbook: Show what one playbook does, what it needs from the caller, and the steps it will run. - plan_playbook: Return a costed plan for a playbook: every step, the model behind it and the exact credit cost. - execute_playbook: Run a playbook end to end as the signed-in user.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "playbooks",
"title": "Playbooks",
"description": "Playbooks tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_playbooks: List Draftless playbooks — named, ready-made workflows (Ad Strategist, Hook Multiplier, Effect Recreate…) grouped into bundles.\n- get_playbook: Show what one playbook does, what it needs from the caller, and the steps it will run.\n- plan_playbook: Return a costed plan for a playbook: every step, the model behind it and the exact credit cost.\n- execute_playbook: Run a playbook end to end as the signed-in user.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_playbooks",
"get_playbook",
"plan_playbook",
"execute_playbook"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Shorts tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - create_shorts_preset: Save a named, reusable creative formula — aspect ratio, model, look/effect, caption style, target platform, brand kit, character, and voice - list_shorts_presets: List the creative presets you can use — your own plus any shared into an organization you belong to. - update_shorts_preset: Edit a saved creative preset in place — only the fields you pass change. - delete_shorts_preset: Delete a saved creative preset. - apply_shorts_preset: Turn a saved creative preset into a ready-to-run payload — merged inputs (model, aspect ratio, duration, negative prompt, voice) plus the ex - create_shorts_session: Create a durable, resumable multi-shot shorts run from a saved preset (or ad-hoc settings). - list_shorts_sessions: List your shorts sessions — id, title, status, and shot progress — newest first. - resume_shorts_session: Start or resume a shorts session, generating only the shots that are not already done. - list_shorts_styles: List the style looks that create_shorts_restyle can apply to a source video. - create_shorts_restyle: Restyle one source video (4-120 s) into styled short pieces with a look from list_shorts_styles (or your own style_image_url).
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "shorts",
"title": "Shorts",
"description": "Shorts tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- create_shorts_preset: Save a named, reusable creative formula — aspect ratio, model, look/effect, caption style, target platform, brand kit, character, and voice \n- list_shorts_presets: List the creative presets you can use — your own plus any shared into an organization you belong to.\n- update_shorts_preset: Edit a saved creative preset in place — only the fields you pass change.\n- delete_shorts_preset: Delete a saved creative preset.\n- apply_shorts_preset: Turn a saved creative preset into a ready-to-run payload — merged inputs (model, aspect ratio, duration, negative prompt, voice) plus the ex\n- create_shorts_session: Create a durable, resumable multi-shot shorts run from a saved preset (or ad-hoc settings).\n- list_shorts_sessions: List your shorts sessions — id, title, status, and shot progress — newest first.\n- resume_shorts_session: Start or resume a shorts session, generating only the shots that are not already done.\n- list_shorts_styles: List the style looks that create_shorts_restyle can apply to a source video.\n- create_shorts_restyle: Restyle one source video (4-120 s) into styled short pieces with a look from list_shorts_styles (or your own style_image_url).",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"create_shorts_preset",
"list_shorts_presets",
"update_shorts_preset",
"delete_shorts_preset",
"apply_shorts_preset",
"create_shorts_session",
"list_shorts_sessions",
"resume_shorts_session",
"list_shorts_styles",
"create_shorts_restyle"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": true
}
}Workflows tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_workflows: List Draftless workflows — named, runnable pipelines you can inspect and then execute with `execute_workflow`. - execute_workflow: Execute a published Draftless workflow by slug. - list_recent_runs: List the signed-in user's most recent Draftless workflow executions with status, duration and credits used. - quote_workflow: The exact credits one run of a workflow costs, line by line, from the same pricing the charge uses. - create_schedule: Run a workflow on a timer or by signed webhook. - list_schedules: List your schedules and your workspaces' schedules: type, timing, whether it is active or paused (and why), the next run, and the per-run an - list_schedule_runs: The run history of one schedule: status (completed, failed, skipped_over_cap), what each run was quoted and charged, and the error for a fai - pause_schedule: Pause a schedule so it stops running, or pass resume:true to start it again. - delete_schedule: Permanently delete a schedule and its run history.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "workflows",
"title": "Workflows",
"description": "Workflows tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_workflows: List Draftless workflows — named, runnable pipelines you can inspect and then execute with `execute_workflow`.\n- execute_workflow: Execute a published Draftless workflow by slug.\n- list_recent_runs: List the signed-in user's most recent Draftless workflow executions with status, duration and credits used.\n- quote_workflow: The exact credits one run of a workflow costs, line by line, from the same pricing the charge uses.\n- create_schedule: Run a workflow on a timer or by signed webhook.\n- list_schedules: List your schedules and your workspaces' schedules: type, timing, whether it is active or paused (and why), the next run, and the per-run an\n- list_schedule_runs: The run history of one schedule: status (completed, failed, skipped_over_cap), what each run was quoted and charged, and the error for a fai\n- pause_schedule: Pause a schedule so it stops running, or pass resume:true to start it again.\n- delete_schedule: Permanently delete a schedule and its run history.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_workflows",
"execute_workflow",
"list_recent_runs",
"quote_workflow",
"create_schedule",
"list_schedules",
"list_schedule_runs",
"pause_schedule",
"delete_schedule"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": true
}
}Commerce tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_store_connections: List the e-commerce / CMS store connections (Shopify, Webflow, …) the signed-in user can access, with their id, provider, display name, and - list_catalog: Read the products (Shopify) or CMS items (Webflow) from a store connection. - push_to_store: Write a single value back to one catalog item — e.g. - enrich_catalog: End-to-end: read a store's catalog, AI-generate an image per item, and push it back (defaults to items missing images).
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "commerce",
"title": "Commerce",
"description": "Commerce tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_store_connections: List the e-commerce / CMS store connections (Shopify, Webflow, …) the signed-in user can access, with their id, provider, display name, and \n- list_catalog: Read the products (Shopify) or CMS items (Webflow) from a store connection.\n- push_to_store: Write a single value back to one catalog item — e.g.\n- enrich_catalog: End-to-end: read a store's catalog, AI-generate an image per item, and push it back (defaults to items missing images).",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_store_connections",
"list_catalog",
"push_to_store",
"enrich_catalog"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": true
}
}Social tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_social_accounts: List the social accounts connected to Draftless — TikTok, Instagram Business, Facebook Pages, and Meta ad accounts — with the connection id - prepare_publish: Step 1 of publishing to TikTok. - publish_post: Publish a video or image PUBLICLY to a connected TikTok, Instagram Business, or Facebook Page account. - get_publish_status: Check whether a publish_post job has gone live. - create_ad_draft: Turn a generated image or video into a Meta ad. - get_ad_performance: Read what the user's Meta ads actually did over a recent window — spend, results, cost per result, click-through and video hook retention pe - record_social_outcome: Record the observed performance of a short-form post (impressions, engagement rate, watch time, CTR).
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "social",
"title": "Social",
"description": "Social tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_social_accounts: List the social accounts connected to Draftless — TikTok, Instagram Business, Facebook Pages, and Meta ad accounts — with the connection id \n- prepare_publish: Step 1 of publishing to TikTok.\n- publish_post: Publish a video or image PUBLICLY to a connected TikTok, Instagram Business, or Facebook Page account.\n- get_publish_status: Check whether a publish_post job has gone live.\n- create_ad_draft: Turn a generated image or video into a Meta ad.\n- get_ad_performance: Read what the user's Meta ads actually did over a recent window — spend, results, cost per result, click-through and video hook retention pe\n- record_social_outcome: Record the observed performance of a short-form post (impressions, engagement rate, watch time, CTR).",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_social_accounts",
"prepare_publish",
"publish_post",
"get_publish_status",
"create_ad_draft",
"get_ad_performance",
"record_social_outcome"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": true
}
}Shots tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - create_shot_bundle: Start a shot bundle: the control-pass package a Blender blockout exports (beauty, depth, normals, optical flow, object masks, first frame, c - finalize_shot_bundle: Confirm a shot bundle's uploads landed. - list_shots: List the Blender shot bundles the caller has uploaded, with their passes, length, dimensions, whether they are complete, and the credit pric - get_shot_bundle: Get short-lived signed read URLs for one shot bundle's passes, so the blockout can be inspected (or handed to another tool) before any credi - generate_shot_video: Turn a Blender shot bundle into a finished video. - get_shot_video_status: Poll a generate_shot_video job.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "shots",
"title": "Shots",
"description": "Shots tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- create_shot_bundle: Start a shot bundle: the control-pass package a Blender blockout exports (beauty, depth, normals, optical flow, object masks, first frame, c\n- finalize_shot_bundle: Confirm a shot bundle's uploads landed.\n- list_shots: List the Blender shot bundles the caller has uploaded, with their passes, length, dimensions, whether they are complete, and the credit pric\n- get_shot_bundle: Get short-lived signed read URLs for one shot bundle's passes, so the blockout can be inspected (or handed to another tool) before any credi\n- generate_shot_video: Turn a Blender shot bundle into a finished video.\n- get_shot_video_status: Poll a generate_shot_video job.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"create_shot_bundle",
"finalize_shot_bundle",
"list_shots",
"get_shot_bundle",
"generate_shot_video",
"get_shot_video_status"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Billing tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - show_plans_and_credits: Billing overview for the signed-in user: current credit balance, plan tier, lifetime earned/spent, live subscription state (package, interva - get_credit_balance: Get the signed-in user's Draftless credit balance, plan tier, and lifetime earned/spent totals. - list_transactions: Paginated credit ledger for the signed-in user: every credit movement, newest first, with amount (positive = earned/purchased, negative = sp - list_invoices: List the signed-in user's Stripe invoices (most recent first) with date, amount in the currency's minor unit, status, line description, and - start_credit_purchase: Create a Stripe Checkout link for a one-off credit top-up and return the URL. - get_purchase_status: Check whether a credit purchase started by start_credit_purchase has actually completed. - get_billing_portal_url: Return a link to the signed-in user's Stripe billing portal, where THEY can change plan, cancel, update payment methods, and download invoic - get_auto_refill_settings: View the signed-in user's automatic credit top-up (auto-refill) configuration: whether it is on, the balance threshold that triggers it, whi - set_auto_refill_settings: Reduce or disable the signed-in user's automatic credit top-up.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "billing",
"title": "Billing",
"description": "Billing tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- show_plans_and_credits: Billing overview for the signed-in user: current credit balance, plan tier, lifetime earned/spent, live subscription state (package, interva\n- get_credit_balance: Get the signed-in user's Draftless credit balance, plan tier, and lifetime earned/spent totals.\n- list_transactions: Paginated credit ledger for the signed-in user: every credit movement, newest first, with amount (positive = earned/purchased, negative = sp\n- list_invoices: List the signed-in user's Stripe invoices (most recent first) with date, amount in the currency's minor unit, status, line description, and \n- start_credit_purchase: Create a Stripe Checkout link for a one-off credit top-up and return the URL.\n- get_purchase_status: Check whether a credit purchase started by start_credit_purchase has actually completed.\n- get_billing_portal_url: Return a link to the signed-in user's Stripe billing portal, where THEY can change plan, cancel, update payment methods, and download invoic\n- get_auto_refill_settings: View the signed-in user's automatic credit top-up (auto-refill) configuration: whether it is on, the balance threshold that triggers it, whi\n- set_auto_refill_settings: Reduce or disable the signed-in user's automatic credit top-up.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"show_plans_and_credits",
"get_credit_balance",
"list_transactions",
"list_invoices",
"start_credit_purchase",
"get_purchase_status",
"get_billing_portal_url",
"get_auto_refill_settings",
"set_auto_refill_settings"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": true
}
}Workspace tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_workspaces: List every Draftless workspace (organization) the signed-in user belongs to, with their role in each (owner / admin / member / viewer) and w - select_workspace: Set the active Draftless workspace (organization) for this MCP connector. - get_workspace: Return detail for the active Draftless workspace (organization): name, your role, plan / subscription status, member count with a role break - list_audit_events: List the audit trail of a shared workspace, newest first: members added or removed, role changes, approval requests and decisions, and chang - get_webhook_secret: Get (auto-provisioning on first call) the HMAC signing secret Draftless uses to sign completion webhooks.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "workspace",
"title": "Workspace",
"description": "Workspace tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_workspaces: List every Draftless workspace (organization) the signed-in user belongs to, with their role in each (owner / admin / member / viewer) and w\n- select_workspace: Set the active Draftless workspace (organization) for this MCP connector.\n- get_workspace: Return detail for the active Draftless workspace (organization): name, your role, plan / subscription status, member count with a role break\n- list_audit_events: List the audit trail of a shared workspace, newest first: members added or removed, role changes, approval requests and decisions, and chang\n- get_webhook_secret: Get (auto-provisioning on first call) the HMAC signing secret Draftless uses to sign completion webhooks.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_workspaces",
"select_workspace",
"get_workspace",
"list_audit_events",
"get_webhook_secret"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Remotion tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - create_remotion_composition: Upload a finalized Remotion project ONCE as a zip of its source: the registerRoot entry point and its files, no node_modules. - update_remotion_composition: Step 2 of the large-zip path: after PUTting the zip to the uploadUrl from create_remotion_composition, call this with upload_complete:true t - list_remotion_compositions: List the caller’s registered Remotion compositions with their ids, entry points and stored default props. - create_remotion_render: Render a registered composition to MP4 (or a single PNG still) on Draftless's render worker, with new input props — iterate without the sour - get_remotion_render: Status of a create_remotion_render job.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "remotion",
"title": "Remotion",
"description": "Remotion tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- create_remotion_composition: Upload a finalized Remotion project ONCE as a zip of its source: the registerRoot entry point and its files, no node_modules.\n- update_remotion_composition: Step 2 of the large-zip path: after PUTting the zip to the uploadUrl from create_remotion_composition, call this with upload_complete:true t\n- list_remotion_compositions: List the caller’s registered Remotion compositions with their ids, entry points and stored default props.\n- create_remotion_render: Render a registered composition to MP4 (or a single PNG still) on Draftless's render worker, with new input props — iterate without the sour\n- get_remotion_render: Status of a create_remotion_render job.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"create_remotion_composition",
"update_remotion_composition",
"list_remotion_compositions",
"create_remotion_render",
"get_remotion_render"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Scene tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_scene_projects: List the caller's 3D scene projects (newest first) with revision and preview. - get_scene_project: Read one scene project: revision, scene sequence (pass both back as guards), the typed scene document and 1 h signed GLB and preview URLs. - create_scene_project: Create an empty 3D scene project. - execute_scene_query: Inspect a 3D scene and render preview images; the scene itself is not changed. - execute_scene_edit: Commit one coherent edit to a 3D scene from a prompt (typed scene operations) or explicit ops. - execute_scene_tool: Run one deterministic scene tool: import (10 cr), rig (16), animate (12) or render (6 + 4/s, doubled at 1080p, up to 20 s). - get_scene_operation: Poll a scene operation: status, logs tail, signed artifacts and revisionAfter. - upload_scene_asset: Mint a one-shot signed upload URL for a .glb, .gltf, .fbx, .obj asset (up to 64 MB). - add_generated_mesh: Copy a finished generate_3d mesh (a completed three_d job) into the scene asset folder, server-side, and return the assetPath for execute_sc - list_scene_assets: List the caller's importable meshes: uploads already in the scene asset folder and finished generate_3d jobs. - get_scene_export: Sign a download for the scene: glb, blend or mp4 (1 h link), optionally at a past revision. - generate_scene_angles: Camera angles of a scene as photoreal stills: 1-6 orbit renders on the worker, then a Seedream 5 Pro edit of each at 1k, 1.5k, 2k in one of - generate_scene_angle_retry: Re-run ONE angle of a generate_scene_angles job whose image edit failed and was refunded (get_scene_operation lists it with retryable:true). - create_scene_shot: Turn the scene into a video shot. - get_scene_view: Render-ready view of a scene for a widget or a quick look: project, optional operation, artifacts and a web link to the Scene Builder. - list_scene_catalog: Search the curated CC0 model catalog (Poly Haven: furniture, props, plants, containers, lighting, themed kits like the_shed or hidden_alley) - add_scene_catalog_asset: Bring one catalog model (from list_scene_catalog) into a scene as a named, stageable object.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "scene",
"title": "Scene",
"description": "Scene tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_scene_projects: List the caller's 3D scene projects (newest first) with revision and preview.\n- get_scene_project: Read one scene project: revision, scene sequence (pass both back as guards), the typed scene document and 1 h signed GLB and preview URLs.\n- create_scene_project: Create an empty 3D scene project.\n- execute_scene_query: Inspect a 3D scene and render preview images; the scene itself is not changed.\n- execute_scene_edit: Commit one coherent edit to a 3D scene from a prompt (typed scene operations) or explicit ops.\n- execute_scene_tool: Run one deterministic scene tool: import (10 cr), rig (16), animate (12) or render (6 + 4/s, doubled at 1080p, up to 20 s).\n- get_scene_operation: Poll a scene operation: status, logs tail, signed artifacts and revisionAfter.\n- upload_scene_asset: Mint a one-shot signed upload URL for a .glb, .gltf, .fbx, .obj asset (up to 64 MB).\n- add_generated_mesh: Copy a finished generate_3d mesh (a completed three_d job) into the scene asset folder, server-side, and return the assetPath for execute_sc\n- list_scene_assets: List the caller's importable meshes: uploads already in the scene asset folder and finished generate_3d jobs.\n- get_scene_export: Sign a download for the scene: glb, blend or mp4 (1 h link), optionally at a past revision.\n- generate_scene_angles: Camera angles of a scene as photoreal stills: 1-6 orbit renders on the worker, then a Seedream 5 Pro edit of each at 1k, 1.5k, 2k in one of \n- generate_scene_angle_retry: Re-run ONE angle of a generate_scene_angles job whose image edit failed and was refunded (get_scene_operation lists it with retryable:true).\n- create_scene_shot: Turn the scene into a video shot.\n- get_scene_view: Render-ready view of a scene for a widget or a quick look: project, optional operation, artifacts and a web link to the Scene Builder.\n- list_scene_catalog: Search the curated CC0 model catalog (Poly Haven: furniture, props, plants, containers, lighting, themed kits like the_shed or hidden_alley)\n- add_scene_catalog_asset: Bring one catalog model (from list_scene_catalog) into a scene as a named, stageable object.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_scene_projects",
"get_scene_project",
"create_scene_project",
"execute_scene_query",
"execute_scene_edit",
"execute_scene_tool",
"get_scene_operation",
"upload_scene_asset",
"add_generated_mesh",
"list_scene_assets",
"get_scene_export",
"generate_scene_angles",
"generate_scene_angle_retry",
"create_scene_shot",
"get_scene_view",
"list_scene_catalog",
"add_scene_catalog_asset"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}Projects tools for the Draftless AI creative studio. Requires sign-in to run. Actions: - list_folders: List Asset Library folders (projects) you can see, scoped to the active workspace when one is selected. - create_folder: Create a library folder (project), optionally inside parent_id. - update_folder: Rename a library folder (name) and/or move it under another folder (parent_id; null = top level). - delete_folder: Delete a library folder AND its subfolders. - update_asset_folder: File up to 50 Asset Library items into a folder (folder_id), or unfile them (folder_id null). - list_explore: Browse the public Explore feed: screened community creations (filter by kind, studio or contest slug) or, with type:'projects', published pr - list_contests: The weekly creator contests with their status from the database clock (upcoming / open / judging / closed), allowed kinds (image, video), cl - create_contest_entry: Enter one of YOUR library images in an open contest: it is shared publicly to Explore, tagged with the contest. - get_project_page: Whether a library folder is published as a public project page (/p/<slug>): slug, title, published/hidden state, item count and URL. - update_project_page: Change the title and/or brief of your public project page (/p/<slug>). - unpublish_project: Take your public project page (/p/<slug>) offline.
| Parameter | Type | Description |
|---|---|---|
| input | object | Arguments for the action (see discover.describe_action once signed in). |
| action * | string | Which operation to run. |
{
"name": "projects",
"title": "Projects",
"description": "Projects tools for the Draftless AI creative studio. Requires sign-in to run. Actions:\n- list_folders: List Asset Library folders (projects) you can see, scoped to the active workspace when one is selected.\n- create_folder: Create a library folder (project), optionally inside parent_id.\n- update_folder: Rename a library folder (name) and/or move it under another folder (parent_id; null = top level).\n- delete_folder: Delete a library folder AND its subfolders.\n- update_asset_folder: File up to 50 Asset Library items into a folder (folder_id), or unfile them (folder_id null).\n- list_explore: Browse the public Explore feed: screened community creations (filter by kind, studio or contest slug) or, with type:'projects', published pr\n- list_contests: The weekly creator contests with their status from the database clock (upcoming / open / judging / closed), allowed kinds (image, video), cl\n- create_contest_entry: Enter one of YOUR library images in an open contest: it is shared publicly to Explore, tagged with the contest.\n- get_project_page: Whether a library folder is published as a public project page (/p/<slug>): slug, title, published/hidden state, item count and URL.\n- update_project_page: Change the title and/or brief of your public project page (/p/<slug>).\n- unpublish_project: Take your public project page (/p/<slug>) offline.",
"inputSchema": {
"type": "object",
"required": [
"action"
],
"properties": {
"input": {
"type": "object",
"description": "Arguments for the action (see discover.describe_action once signed in)."
},
"action": {
"enum": [
"list_folders",
"create_folder",
"update_folder",
"delete_folder",
"update_asset_folder",
"list_explore",
"list_contests",
"create_contest_entry",
"get_project_page",
"update_project_page",
"unpublish_project"
],
"type": "string",
"description": "Which operation to run."
}
},
"additionalProperties": false
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": true
}
}Public, read-only listing of the Draftless tools. To run them, connect https://mcp.draftless.dev/mcp and sign in with a Draftless account (new accounts get free starter credits). Paid actions always show a price preview first.
Names, descriptions and schemas were written by the server and are shown as data.