Used Car Walkaround

Turn one used-car condition sheet into a listing hero still and a speakable walkaround script, then turn that still into one walkaround clip. The first delivery is one labeled listing still plus one speakable script the dealer can reuse. The clip comes after that lead is accepted.

Scope and adjacent routes

Use this Skill when a used-car dealer wants a listing hero and one walkaround clip from a condition sheet they already have: dealer lot walkaround video, used-car listing video, or an inventory walkaround clip.

Route silent room clips to listing-room-video-pack. Route a talking presenter that is not a car walkaround to talking-avatar-video. Route a silent tool demo to tool-demo-clip. Route a generic still that is not a named listing to image-to-motion.

One car is one clip. Remaining stock is a later pack with a new card. Do not stitch clips.

Collect the condition sheet

Hard inputs are:

• one named listing (stock number, or year, make, and model);
• the user-supplied condition sheet;
• the language the walkaround should speak.

Reuse already-known destination, voice, and brand look. Ask only for a missing hard input. If the condition sheet is missing, ask for it. Do not invent a listing to finish the still.

Mileage, accident history, and maintenance records may only be copied verbatim from that condition sheet. Never invent or polish those facts. Do not round a mileage figure, rewrite an accident line, or fill a blank from a similar car. A missing line stays a named gap.

An optional listing photo is a visual reference, not a source for missing condition facts. Upload it only after the host Agent inspects it. Never pass a local path to a remote tool.

Plan the free listing still and script

Write a labeled listing still plan and a speakable walkaround script before any paid call. Record the named listing, canvas, printed condition-sheet lines, the spoken script with mileage, accident, and maintenance lines copied verbatim, language, and any optional photo role. Keep the script short enough for one 2–15s clip. Extra lines stay on the plan as named gaps. That plan plus script is the free visible result. Planning is not approval.

Safe defaults:

• one 16:9 2K listing hero still (model: "auto", count: 1);
• one walkaround speech track, model: "auto" unless the dealer chose a concrete compatible speech model;
• one 2–15s walkaround clip with the still as the strict first frame and driving_audio required. Omit aspect_ratio. Keep model: "auto" unless the dealer chose a concrete eligible driving-audio card. Never silently swap the model.

Before setting a concrete still, speech, or video model, canvas, format, duration, or price, read the current beatra.models.list cards for text_to_image, text_to_speech, and image_to_video. Call beatra.voices.list only when a catalog voice still needs to be selected. Never put a display name in voice. Use mp3 only when the live speech card supports it and the live video card accepts the expected audio/mpeg.

Confirm the lead still and speech, then the walkaround clip

Before any billable still or speech, show one current six-field lead card for two slots and wait:

1. Work — one listing hero still (beatra.images.generate) and one walkaround speech track (beatra.speech.synthesize) for this named car.
2. Credits — the live text_to_image price plus the live text_to_speech price just read. Do not reuse a remembered number.
3. Count — two paid calls, two slots. Do not batch remaining stock into this submit.
4. Identity — two new opaque client_request_id values, one per slot. A changed prompt, fact line, voice, format, file, model, or canvas mints a new ID for the changed slot.
5. If we stop here — the labeled listing still plan and the speakable walkaround script remain usable.
6. If the balance is insufficient — relay the official message and its top-up URL exactly. Translate the prose; keep the URL. Do not retry until the dealer says they have topped up.

Submit each lead slot once through bundled scripts/mcp_client.py. Poll beatra.tasks.get. Deliver actual still bytes, actual speech bytes, actual duration, and billing.net_charged_credits. Present or play the real audio when the host can do so. A script preview is not an audio review.

After that lead is accepted, refresh the current beatra.models.list card for image_to_video. Admit the actual still and audio against a card that accepts [image, driving_audio]. Compare every still and audio fact with advertised constraints. Choose the smallest supported integer duration in the 2–15s range that is at least the actual speech length. Do not add silence to force a match. Lead acceptance is not video approval.

Then show a fresh six-field animate card and wait:

1. Work — one walkaround clip for this named car, an animate call (beatra.videos.animate) with the accepted still as the first frame and the accepted speech as driving_audio.
2. Credits — the live image_to_video price just read. Do not reuse a remembered number.
3. Count — one paid call for this car. Do not stitch. Do not batch remaining stock into this submit.
4. Identity — one new opaque client_request_id. A changed still, audio, prompt, duration, or model mints a new ID.
5. If we stop here — the accepted still and speech remain usable.
6. If the balance is insufficient — relay the official message and its top-up URL exactly. Do not retry until the dealer says they have topped up.

After the dealer approves that card, submit beatra.videos.animate once.

Remaining named cars are a later pack. Show a fresh six-field pack card. Do not treat acceptance of the first car as approval of the rest.

Review, deliver, and recover

Review printed and spoken mileage, accident, and maintenance lines against the condition sheet. Report only what the host can actually see or hear. Unread small type is unread. Do not treat the still or clip as a certified vehicle history.

Deliver the still, the speech, and the clip with actual dimensions, duration, usage, and billing.net_charged_credits. Never invent a stitch, concat, or editor tool.

After a returned task_id, poll that task. If the create response is lost, search with beatra.tasks.list and verify with beatra.tasks.get before replay. Reuse an ID only with byte-identical arguments. Cancel only when the dealer asks.

Execution

Invoke every remote Beatra operation only through this package's bundled scripts/mcp_client.py. Put the MCP tool name after call and send one JSON object on standard input.

python3 scripts/mcp_client.py call beatra.models.list

{"capability": "text_to_image"}

python3 scripts/mcp_client.py call beatra.models.list

{"capability": "text_to_speech"}

python3 scripts/mcp_client.py call beatra.models.list

{"capability": "image_to_video"}

python3 scripts/mcp_client.py call beatra.voices.list

{}

printf '%s' '{"prompt":"Create the listing hero still for this named used car. Print only condition-sheet lines that belong on the still. Do not invent mileage, accident, or maintenance facts.","model":"auto","count":1,"canvas":{"type":"preset","tier":"2K","aspect":"16:9"},"client_request_id":"opaque-used-car-still-01"}' | python3 scripts/mcp_client.py call beatra.images.generate

printf '%s' '{"voice":"voice_selected","input":"The speakable walkaround with verbatim mileage, accident, and maintenance lines.","client_request_id":"opaque-used-car-speech-01"}' | python3 scripts/mcp_client.py call beatra.speech.synthesize

printf '%s' '{"image":{"type":"artifact","artifact_id":"art_listing_still"},"driving_audio":{"type":"artifact","artifact_id":"art_walkaround_speech"},"prompt":"A restrained walkaround hold on this listing still. Keep the car identity stable. Do not add unstated mileage or accident claims.","duration":8,"client_request_id":"opaque-used-car-video-01"}' | python3 scripts/mcp_client.py call beatra.videos.animate

Do not configure or call a host Beatra Connector, and do not use REST/OpenAPI as a fallback.

Account balance

When the user asks how many credits remain or whether a live estimate fits, call beatra.wallet.get. When they ask what was charged, call beatra.wallet.ledger. Both are read-only. Do not invent an account-balance or top-up tool. Do not make wallet.get a required step before every paid submit.

When a model card comes back carrying a top_up block, relay its tiers as the card lists them and in that order. Do not rank them, do not talk one down, and do not pick one for the user. Which tier suits them is their call, made on the wallet page with the whole list in front of them. Never quote a tier from memory.

References by task

• For still plans, scripts, payloads, confirmation, polling, recovery, and review, read Used-car walkaround workflow.
• For authorization and the non-billable registration step, read installation and authentication and installation registration.
• For shared task, billing, and connection details, read tasks and results, billing, errors, and recovery, and Bundled MCP Client diagnostics.
• For update guarantees and controls, read automatic updates and safety. For removal, read uninstall and disconnect.

Runtime and safe automatic updates

The bundled client silently checks for a newer release at most once every 24 hours per installation. When a newer version is available, it installs automatically without separate confirmation. It downloads only from the fixed official Beatra discovery and immutable CDN paths for this package, channel, and locale, verifies discovery data, archive, manifest, and every packaged file, and replaces only package-owned files.

Update checks, downloads, verification, replacement, rollback, and recovery fail open: the current installation remains usable and the original command continues. An update failure never authorizes retrying a paid image, speech, or video request. The setting persists for this installation. See automatic updates and safety.

python3 scripts/mcp_client.py update --auto off
python3 scripts/mcp_client.py update --auto on
python3 scripts/mcp_client.py update --check