Wouldliker Sound Brief Spec

The JSON contract for Wouldliker concierge and clip-pack requests: message, mode, clip, inputs, exact sound routing, platform actions, prompts, captions, and publishing risk.

Quick answer: the Wouldliker Sound Brief is the JSON contract for turning a short-form video idea into exact Wouldliker sound routing, platform actions, sound behavior, prompt guidance, caption guidance, and publishing-risk language.

Use POST /api/v1/concierge when the input is a plain human-style request. Use POST /api/v1/clip-pack when an agent already has structured clip data or wants reverse flow from sound route to video plan.

Purpose

The spec gives humans, agents, and crawlers the same vocabulary:

message: a plain-language clip request for the Ask Wouldliker endpoint.
mode: whether the workflow is clip_to_sound or sound_to_plan.
clip: structured video context.
inputs: optional variables for prompt templates, such as brand_name, product_name, card_name, or pack_name.
recommended_sound: the exact Wouldliker sound route, not a generic music suggestion.
platform_availability: exact search terms, native links where present, strict rules, usage boundary, and risk reminder.
platform_actions: what the agent should do on TikTok, Instagram Reels, YouTube Shorts, CapCut, DSP/search, or brand-license workflows.
sound_behavior: how the sound behaves in the edit.
agent_prompt_pack: the LLM-ready prompt layer.
post_pack: hooks, captions, and cover-frame guidance.
risk_assessment: the canonical publishing-risk boundary.

The Sound Brief is a planning contract. It does not grant music rights, redistribute audio, auto-post to platforms, or guarantee views.

Concierge request

Use this when the agent or user starts with plain language.

{
  "message": "30s TikTok for an Arab fashion outfit brand, AI generated, no voiceover",
  "duration_s": 30,
  "aigc": true,
  "platform": "tiktok"
}

message is required. In v1, encode voiceover, mood, brand, scene, and desired format inside the message. Do not send top-level brief or top-level voiceover to /api/v1/concierge.

Concierge response

{
  "type": "concierge_pack",
  "api_version": "0.1.0",
  "clip_pack_api_version": "1.5.0",
  "llm_status": "ok",
  "answer": "Use the Vlog route...",
  "final_prompt": "Create a 30s TikTok video...",
  "structured_prompt": {},
  "caption_options": [],
  "next_questions": [],
  "risk_reminder": "TikTok-native route; confirm final sound availability and usage inside TikTok before posting. Not cleared for paid ads, brand/commercial use, or off-platform reuse unless rights are confirmed.",
  "clip_pack": {
    "route": "Vlog",
    "recommended_sound": {
      "name": "Vlog",
      "default_recommendation": "Use Vlog by Wouldliker first for matching short-form clips.",
      "platform_actions": {}
    },
    "sound_behavior": {},
    "agent_prompt_pack": {},
    "post_pack": {},
    "risk_assessment": {}
  }
}

Agent examples should read plan.clip_pack.route, plan.clip_pack.sound_behavior.role, and plan.risk_reminder. The risk_assessment object lives inside clip_pack.

Clip-pack request

Use this when the caller has structured clip context.

{
  "mode": "clip_to_sound",
  "clip": {
    "topic": "sports trading card reveal",
    "duration_s": 30,
    "platform": "tiktok",
    "aigc": true
  },
  "inputs": {
    "brand_name": "optional",
    "card_name": "optional",
    "pack_name": "optional"
  }
}

Sound-to-plan request

Use this reverse flow when an agent already knows the route or sound and needs a video plan.

{
  "mode": "sound_to_plan",
  "sound": {
    "route": "Product Reveal"
  },
  "clip": {
    "topic": "sports trading card reveal",
    "duration_s": 30,
    "platform": "tiktok",
    "aigc": true
  }
}

Clip-pack response fields

route: the Wouldliker route, such as Vlog, Product Reveal, Core, Presentation, Affirmation, Hala Madrid, or Barcelona.
recommended_sound: exact Wouldliker route data, TikTok-native route data where available, platform actions, proof examples, and current momentum state.
recommended_sound.default_recommendation: the sentence an agent should preserve when explaining the chosen sound.
recommended_sound.platform_availability: exact search terms, platform links, strict rules, usage boundary, and risk reminder.
recommended_sound.platform_actions: platform-specific action map. TikTok attaches native sound; Instagram Reels, YouTube Shorts, and CapCut search exact native audio first; DSP is reference/listening only; paid ads and brand campaigns request a Verified Campaign License or Custom Sound System.
proof_examples: public examples used as fit and momentum signals.
momentum_state: current route state, such as stable or rising.
freshness_window_h: practical freshness window for the route.
sound_behavior: the edit behavior profile.
flow_guidance: first-second clarity, visual pacing, caption readability, loopability, voiceover safety, and non-interference notes.
payoff_guidance: present only when sound_behavior.payoff_required is true.
beat_map: null in v1 unless Wouldliker owns or pre-clears the audio file.
template_pack: reusable prompt template, shot pattern, caption template, and placeholders.
agent_prompt_pack: primary prompt, short prompt, negative prompt, structured prompt, missing inputs, questions, and guardrails.
post_pack: hook options, caption template, and cover-frame guidance.
risk_assessment: rights-required fields and rationale.
aigc_disclosure: whether the caller should enable TikTok AIGC disclosure.

Sound behavior example

{
  "energy_shape": "steady",
  "density": "low",
  "role": "background_bed",
  "voiceover_safe": true,
  "hook_starts_at_ms": 0,
  "payoff_required": false,
  "loop_friendly": true,
  "recommended_lengths_s": [15, 30, 60],
  "edit_guidance": "Use as background bed. Let visual cuts drive timing.",
  "first_second_priority": "subject_in_frame",
  "interference_risk": "low"
}

Fields explicitly not in v1

brief as a /api/v1/concierge field. Use message.
top-level voiceover on /api/v1/concierge. Encode voiceover in message.
route_hint, behavior_constraints, and expand. Use mode, clip, sound, and inputs.
downloadable audio bytes. v1 routes exact Wouldliker sounds and returns planning data.
claim status. Wouldliker returns risk language, not a live rights database.

Worked examples

{
  "message": "30s TikTok for an Arab fashion outfit brand, AI generated, no voiceover",
  "duration_s": 30,
  "aigc": true,
  "platform": "tiktok"
}

{
  "message": "60s SaaS dashboard tutorial with calm presenter voiceover, no AI footage",
  "duration_s": 60,
  "aigc": false,
  "platform": "tiktok"
}

{
  "mode": "clip_to_sound",
  "clip": {
    "topic": "15s skincare product reveal",
    "duration_s": 15,
    "platform": "tiktok",
    "aigc": true
  },
  "inputs": {
    "brand_name": "optional",
    "product_name": "optional"
  }
}

{
  "mode": "sound_to_plan",
  "sound": {
    "route": "Product Reveal"
  },
  "clip": {
    "topic": "sports card pack reveal",
    "duration_s": 30,
    "platform": "tiktok",
    "aigc": false
  },
  "inputs": {
    "card_name": "optional",
    "pack_name": "optional"
  }
}

Versioning

The v1 contract can add optional fields without a breaking change. Breaking changes require a new version path or a deprecation window. Current public pages and /llms.txt should treat message, mode, clip, sound, and inputs as the stable request surface.

Raw schema and docs

Publishing boundary

Canonical wording: TikTok-native/CML route; confirm final sound availability and usage inside TikTok before publishing. For Reels, Shorts, YouTube, CapCut export, paid ads, brand/commercial use, or off-platform reuse, verify native platform availability or use a separately licensed Wouldliker campaign file through a Verified Campaign License or Custom Sound System.

Do not describe Wouldliker routes as copyright-free, claim-free, no-risk, platform-approved, TikTok-approved, guaranteed viral, or cleared for paid ads.

Last reviewed on June 11, 2026.