Uwear

Use this when the user wants a photoshoot and you have no board_id yet. Creates a new, empty Uwear Shoot Board and returns its board_id, which every other Shoot Board tool takes. This client shows no widget: build the shoot with tools (search_uwear_library, modify_shoot_board, propose_brief) and tell the user what is on the board. prompt: one line describing the shoot; expected_types: what it needs.

Use this before modify_shoot_board, propose_brief or update_brief: the user changes the board in the widget at any time. Returns the board's items (Clothing, Models, Outfits, Locations, art directions, files, results) with ids and names, its revision, its linked brief_id, and brief_inputs: the guide, asset facts, automations, cameras, AI models and limits a brief is written from. Changes nothing.

Use this when the user names Clothing, Models, Locations, art directions, files or past results in chat and you need their ids. Takes a short keyword query (2-5 words, or an exact name, SKU or id) and returns typed ids, names and thumbnails. It finds items; it does not recommend them. Changes nothing.

Use this when the user asks in chat to change what is on the open Shoot Board. add puts items on it; set replaces the whole item list (to remove items, send the ones to keep); clear empties one type, or without type starts a new shoot on the same board (its brief is unlinked; photos stay in the library); add_as_view makes from_garment_id's photo the back, side or detail view of garment_id and takes from_garment_id off the board, only after the user agrees the two show one garment. Items are {type, id} from search_uwear_library or read_shoot_board. To reuse an automation, add {type: automations, id: a workflow_id from brief_inputs.automations}: the board pins one and the next brief runs every look through it. Pass expected_revision from read_shoot_board; on revision_conflict, apply the change to the returned board once.

Moves a Shoot Board that is open in a widget to a view (board, brief, automation, results, brief_history). This client shows no widget, so it changes nothing the user sees: read the board, brief or results with read_shoot_board, get_brief or get_generation_results and tell the user.

Use this when the user attaches images in this chat and wants them in Uwear, on the Shoot Board or in a shoot: clothing, a Model, a Location or a reference. Put each attachment in its own field, image_file_1, image_file_2, ... in attachment order, as the chat provides it: ChatGPT passes the file itself, also when you see only a file name or path. Chat attachments only: a web address is refused (ask the user to attach that photo). kinds: {image_file_N: garment|avatar|location|reference} for photos whose kind the user said or is clear (avatar: a person to use as a Model); Uwear sorts the rest. Saves each photo to the library (a repeat is reused), adds it to the board and returns its pile and asset id for propose_brief / update_brief. A photo returned unsure lists two candidates: ask the user and call again with the same files, kinds and sides. sides: {image_file_N: back} for a garment photo that plainly shows the garment's back (you can see it); left out, a photo is its garment's front.

Use this when a garment's photo shows its back but Uwear has it as the front (or the other way round): you saw the photo in this chat, or the user says so. side back makes the photo the garment's back, so back shots run from it; front undoes that. Call it yourself and continue (confirm_brief if the user asked to run): never ask the user to relabel a photo on the board. The board's brief is saved again with its new quote. Deletes nothing.

Use this when the user asks to create, write or save an art direction, or the board's New art direction asks for one: write it from their words, the chat and the board, and save it in the same turn. Saves it to the library and makes it the board's one art direction, replacing the previous one (which stays in the library). Fill every section with concrete, photographable choices. It applies to every look, so it names no garment, outfit, Model or Location (text naming a board item is refused). If the board has a brief, apply it with update_brief direction.art_direction_id. If the result lists problems, fix them all and call again.

Saves a new brief and shows it on the board: the looks to shoot, their direction and their shots. The previous brief is kept (one that ran stays in Brief History). Returns the saved brief, or the problems to fix.

Use this when the user asks in chat to change the board's brief: its looks, Models, outfits, Locations, references, name, shots and steps, AI model or art direction. Send the complete replacement look list with the brief_id it replaces, keeping each unchanged look's key. Shots or steps: recipe with the complete new shot list (recipe_look_keys for some looks only) or recipe_workflow_id. AI model only: image_model / video_model. Art direction only: art_direction_id. Saves the revised brief on the board; nothing runs. If the result lists problems, fix them all and call again; on revision_conflict, read the board and revise its current brief.

Use this when you need a Shoot Board brief by brief_id: its looks, automations, credit quote (quote.total_credits), status and revision. Shows it on the board; spends nothing.

Generates the photos and videos of brief brief_id. Spends credits at the brief's quote. This client may show no approval prompt, so call it only when the user agreed in chat to run the brief: their message asking to generate or run it is that agreement, a request to change the brief is not one; otherwise give the quote and wait for their yes. expected_brief_revision, runs and expected_total_credits must match the latest brief; if they changed, nothing runs and the current brief is returned. The same idempotency_key never runs or charges twice.

Use this to find a past shoot when the user refers to one ("run the same as last week") or asks what ran: the account's briefs, newest first, 10 by default (limit up to 25; before: next_before for older ones; status ran for generated ones only; automation_workflow_id for the runs of one Automation). Each compact row has its brief_id, title, status, ran_at, automations, direction label, image model, output, look count and credits. Then read the one you need with get_brief for its full values. Changes nothing.

Lists the user's Automations, most recently used first, one row each. 10 per page (limit up to 25; cursor: next_cursor for more). Changes nothing.

Use this after confirm_brief or when the user asks how a shoot is going: each look's run of brief_id with its status, credits, results (generation_result_id, kind, url) by step and the error of every step that failed (failures lists them all). Read it again until done is true; while it runs, say so (the board shows progress). Then show the results with get_generation_results. Changes nothing.

Use this to check one generation job by generation_id: its status, credits and, when it failed, a safe error_message. A brief's runs are read with mcp_brief_runs. Changes nothing.

Use this to show results in chat: generation_result_id (from mcp_brief_runs) for one result, generation_id for a job's results, or query, filters and dates to find past ones. Returns image previews to show as they are (never Markdown-embed result URLs); url is the original. Changes nothing.

List user's garments with structured filters or query for hybrid name/SKU/metadata/image-attribute search. body_zone expands to categories and is ORed with category when both are given. With a category or body_zone filter and include_uncategorized=false, uncategorized_count reports garments hidden by the category filter within the same scope. If matches are short and uncategorized_count > 0, call again with include_uncategorized=true and inspect those garments.

Upload Clothing from public HTTP(S) image URLs. Queues a recoverable batch upload and returns an operation immediately; retry with the exact same arguments to read its state (a retry creates no second upload). Photos are stored as uploaded, without background removal; processing_mode is ignored. Prefer assets with {asset_kind: full|detail, asset_view: front|back|side, asset_url}; exactly one full front is required, every full view is unique, and detail views can repeat. No base64, chat file objects, or local paths.

List the user's Models: the reusable people who wear the Clothing (not the AI models; for those, use mcp_list_model_catalog). Supports structured filters or query for hybrid name/metadata/image-attribute search. Set include_image_url=true when the Model's image itself is needed.

Batch upload reusable Models from person photos the user attached in chat. Each photo becomes the upper_body_front identity anchor. Keep names and optional user-given age_range, build, and height_cm in avatars[], then pass top-level image_file_1, image_file_2, etc. in the same order. Traits are never guessed. No base64, no local paths. Do not use ordinary text-to-image results as an upload fallback. Use only when the user explicitly wants a specific or consistent person.

List user's outfits (saved garment combinations). Filter by garment IDs, active status, date range, or query against contained garments.

List reusable location/scene reference images by structured filters or query: IDs, tags, date range, and sort. Use when the user wants to pick a saved background/location reference for a generation.

Create a reusable location/scene reference from an image URL. If no description is supplied, the backend analyzes the image. Use the returned location_id in propose_brief/update_brief.

Create a new tag for organizing assets (garments, models, outfits, files, results, locations). Requires an active company workspace.

Default tool for generation-result QA. Use this whenever the user asks to QA, quality check, validate, review, inspect for defects, approve/reject, or assess already-created generation results. Do not substitute view_image for QA unless the user explicitly asks for a manual visual critique instead of the official Uwear QA pipeline. First identify the numeric generation_result_ids, then call this tool, then call read_generation_result_qa for the same IDs. QA is free. QA is built for scale — validating large batches (hundreds or thousands of results). When the user is iterating on a handful of results one by one, do not queue QA on your own initiative; run it when the user asks for it or when operating at batch scale.

Read official Uwear QA status, decision, and structured QA JSON for existing generation results. Call this after queue_generation_result_qa when the user asks to QA/check/validate/review generated outputs; summarize the structured decision and issues, not a manual view_image opinion.

The result viewer's star. Marks one of the user's own generation results as a favorite (favorite true) or removes the mark (false); deletes nothing and spends no credits.

ONLY for video montage/stitching/export workflows. Use when the user explicitly asks to create a montage, stitch clips, make a reel, export a video sequence, make video clips from images, or combine images/videos into one final video. Never use this for a photoshoot, lookbook, product shoot, collection shoot, outfit shoot, garment shoot, or image-generation request; those must use request_user_context followed by propose_brief/update_brief. Do not call this merely because selected context contains images, generations, garments, or models. A photoshoot may later feed a montage, but the photoshoot itself must be proposed as a BriefProposal first. PROPOSES the montage for user review — user can edit clips, generate missing videos, then export. Supports: existing videos with optional trim (`target_duration` or `start_time`/`end_time`), images that need video generation (specify video_model + a bespoke per-image motion prompt, and optionally `target_duration` or `duration`), per-clip speed/mute, global aspect ratio. If the user asks for clips to be e.g. '3 seconds each', set `target_duration: 3` on every item, including image items. For image items, avoid generic repeated prompts: tailor each prompt to the specific image and any requested zoom, movement, energy, or camera direction. If motion is not specified, inspect the image first with view_image and then write a fitting motion prompt from the image content before proposing. The user reviews and confirms in the UI. Export is free (0 credits); video generation clips cost credits per their model.

Patch the latest MontageProposal already shown in the UI. Use only for an existing video montage/reel/stitched sequence proposal; do not use this to modify a photoshoot brief. For photoshoot brief changes, call update_brief instead. Prefer `operations`, an ordered list of typed montage edits such as `clear_trim`, `remove`, `set_duration`, `set_trim_range`, `set_speed`, `set_muted`, `set_prompt`, `set_video_model`, `set_generate_audio`, `set_last_frame`, `set_aspect_ratio`, and `set_base_resolution`. Use `set_trim_range` for requests like 'use the second half', 'start at 3s', or 'trim from 3s to 6s'. Operations are applied in order to the current montage state. Legacy flat fields are still accepted as a compatibility shim. This updates the proposal only; it does not execute export or generation.

Get user's current credit balance.

Takes no parameters.

List available tags for the user's company. Use this to resolve user-provided tag names like 'summer26' before selecting garments, outfits, models, files, generations, or locations. For no-UI MCP flows, find the tag ID here, then call get_items_by_tag or the relevant list_* tool with tag_ids.

Takes no parameters.

Batch save reference images from chat attachments into the Files library. Keep file metadata in files[] and pass top-level image_file_1, image_file_2, etc. in the same order. No base64, no local paths. Use for mood boards, backgrounds, and other reference images that are not Clothing or Models.

List selectable ArtDirections as compact summaries: visible system ArtDirections plus company ArtDirections. Use query to match exact names and natural-language creative requests. Use this as the catalog step before choosing an art_direction_id for a brief; if there is no clear match, fall back to the user's phrase as prompt/creative_context. Use get_art_direction for the full markdown document before reusing one in a brief.

Get a selectable ArtDirection by ID.

List canonical company templates, optionally filtered by kind, scope, command use case, or public t-xxxx share code. Templates persist canonical commands but do not execute generations. System templates named `Demo — …` are worked examples of the canonical command language; read one with `get_template` before authoring a first brief or workflow.

Read the Uwear Shoot Board.

Replace the Uwear Shoot Board items.

Append items to the Uwear Shoot Board.

Clear all Shoot Board items, or only one context type.

List uploaded (reference) files for the Uwear MCP workspace files tab and the Shoot Board's References tab.

Save an edited image (image_base64) as a new child result of parent_generation_result_id. Changes no existing result.

Export montage clips as one stitched video. Spends credits: call it only after the user confirmed the clips and their order in chat. Returns the generation_id to poll with get_generation_status.

Queue an image-to-video generation for one montage clip. Spends credits: call it only after the user agreed in chat. Returns the generation_id to poll with get_generation_status.

Sort photos dropped on the Shoot Board into garment, avatar, location or reference (or unsure with two candidates) and report duplicates within the drop and in the library. Saves nothing. A photo the call had no time left for comes back with error deferred: send it again.

Save sorted Shoot Board photos to the library as garments, avatars, locations or reference files, skip ones already saved, and add them to the Shoot Board at expected_revision; a board that changed is refused before anything is saved (revision_conflict with the current board).

Move a Shoot Board photo between the Clothing, Models and Locations piles. Saves the same image as the new kind and swaps it on the board; the old asset is deleted only if intake saved it in the last 24 hours and nothing uses it, otherwise it stays in the library. Asking for a move that already happened changes nothing and returns the board (status "unchanged").

A Clothing card dropped on another. The photo of from_garment_id becomes the back, side or detail view of garment_id, and from_garment_id leaves the board (deleted only when it is a fresh drop nothing else uses). Or a photo put in the brief's missing-photo slot: image_url instead of from_garment_id, nothing leaves the board. The board's draft brief is saved again when the view changes it. Returns the board (and brief); asking again answers status unchanged.

Store one photo dropped on the Shoot Board, sent inline as base64, for a host that cannot upload files itself. Returns a receipt (sha256, size, stored, image_url, image_url_expires_at); pass image_url to mcp_intake_files and mcp_commit_intake. Without data, stores nothing and returns the receipt of the photo already stored under sha256 (a fresh image_url), or stored: false.

Edit the Shoot Board's brief by hand: add, remove or change looks, or run looks through another automation. Saves a new brief linked to the board and returns the board, the brief, its quote and the credit delta; dry_run only quotes.

An automation's past runs, newest first: one row per Shoot Board Generate (its looks) or per run started on Uwear, with when, how it went, credits and covers.

The account's Shoot Board briefs, newest first: every brief that ran (on any Shoot Board) and this board's draft, each with its status, looks, automations, credits and covers.

The company's shoot automations (and other automations to copy).

Open the brief's automation in the builder: its draft, or once Generate published it its live version (saving an edit of it makes a new automation); another automation opens read-only.

Validate and price the automation being edited.

Save the automation's draft.

Copy an automation as a new automation for this shoot.

The automation builder reads library items (Clothing, Models, Locations, ...) of one kind by id to show them on its cards. Read-only.

The automation builder's picker: one page of the user's library items of one kind, filtered by search text, tags or category. Read-only.

The account's tags, for the automation builder's tag filters. Read-only.

The garment categories, for the automation builder's category filters. Read-only.

The camera views (front, side, back, crops) a shot or camera edit can use, for the automation builder. Read-only.

The automation format (step kinds and their settings) the automation builder edits against. Read-only.

One generation result as GET /generation-result/{id} returns it (tags, review verdict, QA, prompt, model), for the widget's result viewer.

One generation with its results as GET /generation/{id} returns it (prompt, model, settings, inputs), for the result viewer's Info panel.

The result viewer's review rail. Sets the user's verdict on their own results to approved or rejected, or clears it (decision null), replacing any verdict they had; deletes nothing and spends no credits.

Save a still frame of a video result as a new image result, as POST /generation-result/{id}/extract-frame. The result viewer's Save frame.

Replace a result's tags with tag_ids (empty removes all), as PUT /generation-result/{id}/tags. The result viewer's tag tray. A tag that starts one of the user's automations starts it, as on the platform.

Rename or recolor one of the company's tags, as PATCH /tag/{id}. The result viewer's tag tray.

Delete the user's results and their stored images, as the platform's Delete. Cannot be undone: call it only after the user confirmed in chat which results to delete.

The active AI models with their capabilities and pricing rules, as GET /models returns them, for the widget's model pickers.

Takes no parameters.

Read the user's saved photoshoot defaults (aspect ratio, camera, resolution) with every allowed value. Backs the plugin settings page.

Takes no parameters.

Save changed photoshoot defaults (aspect ratio, camera, resolution). Send only the settings that changed; returns every current value. Defaults never authorize a paid generation.