인터페이스는 현지화되어 있으며, Skill 소개와 설명 문서는 선택한 언어로 AI가 자동 번역합니다. 코드·명령어·경로·전문 용어는 원문을 유지하며 원작자 문서가 우선합니다.
Skills 둘러보기
생산성 및 협업 · SKILL.md

store-pa-broadcast-pack

store-pa-broadcast-pack

Build dated store PA announcement reads for one campaign window: open, promo, flash, and close. This in-store announcement pack and retail PA voice studio records store promo voice lines you can play on the shop floor through the campaign dates. Use it for store opening announcements, flash-sale PA, closing announcements, and in-store promo voiceover.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill 문서

GitHub

name: "store-pa-broadcast-pack" description: "Build dated store PA announcement reads for one campaign window: open, promo, flash, and close. This in-store announcement pack and retail PA voice studio records store promo voice lines you can play on the shop floor through the campaign dates. Use it for store opening announcements, flash-sale PA, closing announcements, and in-store promo voiceover."

Store PA Broadcast Pack

Give a dated campaign four spoken PA reads: open, promo, flash, and close. One store voice. Play them through the campaign window.

Scope and adjacent routes

Use this Skill when a store needs dated in-store PA promo reads for one campaign window.

Route music loops to livestream-bgm-pack. Route live-commerce scripts to live-commerce-script-studio. Do not generate or mix music in this pack.

Inputs and defaults

The hard inputs are:

  • the store or campaign name the floor should hear;
  • the campaign window dates;
  • the offer facts for open, promo, flash, and close, or permission to draft those four reads from the supplied offer;
  • the target language;
  • a pronunciation table whenever brand names, SKUs, or place names appear.

If those names exist and the table is empty, stop and collect the readings. Do not invent an offer or a date window. A speaker sample is optional. Without a sample, choose a catalog voice with beatra.voices.list.

Defaults:

  • four dated reads: open, promo, flash, and close;
  • one store voice for the whole window;
  • language omitted unless the user asks for a language;
  • model: "auto" only when every voice-compatible live card supports the language; format: "mp3"; speed: 1.0;
  • one synthesis call per read so a single read can be redone.

Keep each submitted input at or below 50,000 characters. Split on sentence boundaries, never mid-sentence. Write short spoken sentences. Do not send music cues, ducking notes, or playlist labels into input.

Planning and casting are free. Clone, if used, and each synthesis block are paid.

Clone the store voice only when the user asks. Confirm consent first: the user must state this is their voice or that the speaker authorized this cloning use. Access to a file is not consent. A suitable sample is currently about 10 to 300 seconds, no larger than 20 MiB, and clean single-speaker speech. Only after consent, inspect the sample and upload it through the bundled client. Show the live voice_clone admission card before any clone client_request_id: route voice_clone, tool beatra.voices.clone, live price for one successful activated voice, the provisional estimate, the 600-credit gift usually cannot start this clone, and what happens if the balance is short. Do not offer a free clone. Do not create client_request_id until the user confirms they have topped up or already have enough credits. Submit beatra.voices.clone exactly once:

{
  "sample": {"type": "artifact", "artifact_id": "artifact_..."},
  "display_name": "Store Voice",
  "consent_attested": true,
  "model": "auto",
  "client_request_id": "opaque-store-clone"
}

Poll with beatra.tasks.get until terminal, then freeze that voice_id.

Golden path

Submit every Beatra call through the bundled scripts/mcp_client.py. Do not configure a host Beatra Connector. Do not use REST/OpenAPI as a fallback.

  1. Build the campaign ledger: window dates, slot ID, spoken text, and intended filename for open, promo, flash, and close.
  2. Call beatra.voices.list when a voice is still needed. Keep one voice_id for the whole window. Never put a display name in voice.
  3. Read the live text_to_speech card with beatra.models.list. Confirm the voice, the four reads, weighted-character total, current estimate, and one client_request_id per read before the first paid TTS.
  4. Synthesize each read with beatra.speech.synthesize exactly once. Poll each speech task with beatra.tasks.get until terminal.
  5. Deliver dated labeled clips with slot IDs, plus actual duration, MIME type, URL or artifact, resolved model, and billing.net_charged_credits.
  6. Review that dates, prices, and the store name match the ledger. Redo only the changed read.

Decisions that require confirmation

Confirm the voice and the campaign estimate before the first paid TTS. A changed read is one new paid block.

Recovery

Recover a lost create response only with that read's identical frozen payload and ID. A changed text is a new ID. Recover a lost task ID through beatra.tasks.list and beatra.tasks.get before any resend. Cancel only at the user's request. On 409, keep polling the same task. On insufficient_balance for clone or TTS, keep the top-up URL inside the balance error exact and retry the same frozen ID only after the user says they have topped up.

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

Runtime and safe automatic updates

Use or invoke the bundled scripts/mcp_client.py for every Beatra operation. Before ordinary commands it silently checks for a newer release at most once every 24 hours per installation. Silent checks are enabled by default, and a newer release installs without separate confirmation.

The updater accepts only the fixed official discovery address and immutable Beatra CDN path embedded for this package, channel, and locale. It verifies the discovery data, archive, manifest, and every file's size and checksum before replacement. It replaces only package-owned files and rejects redirects, downgrades, wrong package/channel/locale/version data, unexpected URLs, unsafe archives, and files outside the owned destination.

Update checks, downloads, verification, replacement, rollback, and recovery fail open: the current installation remains usable and the user's original command continues. An update failure never authorizes retrying a paid generation. The automatic-update choice persists across later commands for this installation:

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