# Wouldliker — examples for developers and AI-video tools

Live URL: https://wouldliker.com/developers/examples/

Describe a video or product. Get the recommended Wouldliker TikTok sound, the official TikTok sound URL, public proof, a one-paragraph brief, and a caption pack. Two POST calls. No auth. CORS open.

Public proof is evidence of fit and momentum, not a guarantee of views.

## The two endpoints

- `POST /api/recommend` — returns `primary_sound` (sound_slug, name, page_url, tiktok_sound_url), `why`, `avoid_if`, `saturation_state`, `proof_examples`, `aigc_guidance`, `downgrade_reason`, `copy_boundary`.
- `POST /api/brief` — returns `clip_brief`, `caption_pack.short / mid / with_hashtags`, `hashtag_pack`, `aigc_disclosure_block`.

Hand `tiktok_sound_url` to whatever already attaches sounds at upload time. Wouldliker does not auto-post. Wouldliker does not return audio bytes. Wouldliker does not promise views.

## Available example clients

| Language | Style | File |
| --- | --- | --- |
| curl | Six worked examples (recommend, AIGC, brief, two-call combo, football routing, errors) | https://wouldliker.com/developers/examples/curl/ |
| Node.js | One file, native fetch, `describeToPublishablePack(clip)` helper | https://wouldliker.com/developers/examples/nodejs/wouldliker.js |
| Python | One file, requests-based, `describe_to_publishable_pack(...)` helper | https://wouldliker.com/developers/examples/python/wouldliker.py |
| n8n | Importable workflow, webhook → recommend → brief → shaped output | https://wouldliker.com/developers/examples/n8n/wouldliker-recommend.json |
| Zapier | Two Webhooks-by-Zapier steps, no auth | https://wouldliker.com/developers/examples/zapier-make/ |
| Make.com | Two HTTP modules | https://wouldliker.com/developers/examples/zapier-make/ |
| ComfyUI | Custom node returning SOUND_SLUG, TIKTOK_SOUND_URL, BRIEF, CAPTION_*, COPY_BOUNDARY | https://wouldliker.com/developers/examples/comfyui/wouldliker_node.py |

## Live API endpoints used by the examples

Static GET:

- GET https://wouldliker.com/api/sounds
- GET https://wouldliker.com/api/sounds/{slug}
- GET https://wouldliker.com/api/proof
- GET https://wouldliker.com/api/momentum
- GET https://wouldliker.com/api/openapi.json
- GET https://wouldliker.com/api/health

Dynamic POST:

- POST https://wouldliker.com/api/recommend
- POST https://wouldliker.com/api/brief

## Current routing summary

- Vlog — default route when intent is unclear.
- Product Reveal — ecommerce / product reveals / object showcase / demo.
- Core — harder UGC hook alternative (`ugc`, `hook`, `ad`, `performance`).
- Presentation — talking-head / explainer / caption-led / voice-safe.
- Affirmation — calm beauty / wellness / soft creator.
- Hala Madrid — football payoff / hype / celebration.
- Barcelona — reflective football / elegance / club tribute.

Recommendations are deterministic. Same input returns the same sound.

## Boundaries

- No TikTok auto-posting. The publisher attaches the returned `tiktok_sound_url` at upload time and sets the AIGC label.
- No MP3 / audio redistribution. The API returns links to existing TikTok sound pages, never audio bytes.
- No guaranteed views. Every machine-readable response carries `copy_boundary`.
- Seven sounds, not a stock library. If a clip doesn't fit any of them, the API returns the default route honestly.

## Status

- Phase 1 (static API + JSON data): live.
- Phase 2 (POST /api/recommend, POST /api/brief): live.
- Phase 3 (MCP server, PyPI `wouldliker-mcp`, registry submission): in build, not yet public.

## Contact

DM @wouldliker on TikTok for integration questions.