AniVault Scraper API

The video-sourcing backend for AniVault.

Resolves anime titles → episodes → live, playable streams (sub & dub) across multiple streaming providers, with built-in HLS/subtitle/video proxying so everything plays with clean CORS, no matter what the upstream server sends.

---

How it works

flowchart LR
    Client([AniVault site / Senshi Player]) -->|1: search / info| API[Scraper API]
    API -->|AniList + MAL ID mapping| AniList[(AniList GraphQL)]
    Client -->|2: episodes| API
    Client -->|3: servers| API
    Client -->|4: watch| API
    API -->|scrape| Sources{{Senshi · AnimeHeaven<br/>Miruro · Anikoto}}
    Sources -->|embed / m3u8 / mp4| API
    API -->|proxied stream + subtitles| Client

Loading

A typical playback flow is search → info → episodes → servers → watch: resolve the anime, list its episodes, list which servers have episode N, then resolve that server into an actual playable stream. /watch can also be called directly if you already know the source/episode.

Sources

Source	Status	Notes
Senshi (senshi.live)	✅ Verified	Behind Cloudflare — requires FlareSolverr to solve the challenge before scraping
AnimeHeaven (animeheaven.me)	✅ Verified	Not behind Cloudflare, no FlareSolverr needed — direct MP4 sources
Anikoto (anikoto.net)	✅ Verified	Megacloud/Megaplay decryption + a direct-CDN side channel
Miruro (miruro.tv)	⚠️ Offline	Upstream pipe endpoint is currently down — requests will fail until it comes back

FlareSolverr runs as a separate service; set FLARESOLVERR_URL in the environment so Senshi requests can solve the Cloudflare challenge. The API also pings it periodically to keep it warm on free-tier hosts.

Endpoints

All routes are mounted under /api.

GET /api/search?q=

Search AniList for a title.

GET /api/search?q=naruto
→ { query, count, results[] }

GET /api/info

Resolve an anime's AniList/MAL IDs and per-source site IDs.

Param	Required	Notes
anilistId	one of these
malId	one of these

GET /api/info?malId=20
→ { anilistId, malId, title, siteIds: { zoro, animeheaven, anikoto, ... } }

GET /api/episodes

List episodes for a title on a given source.

Param	Required	Notes
anilistId / malId	yes*	*unless heavenId is used with source=animeheaven
source	no	senshi | animeheaven | miruro | anikoto (default senshi)
heavenId	no	manual AnimeHeaven show id

GET /api/episodes?anilistId=20&source=senshi
→ { anilistId, malId, title, source, siteId, count, episodes[] }

GET /api/servers

List available servers (sub/dub) for a specific episode.

Param	Required	Notes
anilistId / malId	yes*
ep	yes	episode number
type	no	sub | dub | all (default sub)
source	no	default senshi
heavenId	no	manual AnimeHeaven show id

GET /api/servers?anilistId=20&ep=1&type=sub&source=senshi
→ { anilistId, malId, title, episode, type, source, servers[] }

GET /api/watch/:source/:id/:ep/:type

Resolve a real, playable stream for an episode — the main playback endpoint.

Path param	Notes
source	senshi | animeheaven | miruro | anikoto
id	AniList id (or mal-{id}, or AnimeHeaven id)
ep	episode number
type	sub | dub

Query param	Notes
server	prefer a specific server by name
strict	1/true — only use server, don't fall back to others

GET /api/watch/senshi/20/1/sub
GET /api/watch/miruro/20/1/sub?server=bonk-sub
→ { embedUrl, m3u8, hlsProxyUrl, playbackMode, subtitles[], server, availableServers[], ... }

GET /api/watch

Same as above, as query params instead of a path — useful when building a URL dynamically.

GET /api/watch?source=senshi&anilistId=20&ep=1&type=sub

GET /api/proxy/hls?url=&ref=

Proxies an .m3u8 playlist (and rewrites internal segment/key URIs to also route through this proxy) so the browser never hits the upstream CDN directly — fixes CORS and Referer/Origin restrictions.

GET /api/proxy/subtitle?url=&ref=

Proxies a subtitle track with open CORS, converting SRT → WEBVTT on the fly if needed.

GET /api/proxy/video?url=

Proxies a direct MP4 stream (used by AnimeHeaven), forwarding Range requests for seeking.

GET /api/health

→ { status, version, sources[], uptime, cache, timestamp }

Debug (internal)

GET /api/debug/miruro and GET /api/debug/miruro-sources dump the raw Miruro pipe response for a given provider/episode — used for diagnosing the current Miruro outage, not part of the stable public surface.

Tech stack

Runtime	Node.js (Express), TypeScript
Scraping	Cheerio, Axios
Cloudflare bypass	FlareSolverr
Caching	node-cache (in-memory), optional Upstash Redis
Hosting	Railway

Status

Senshi, AnimeHeaven, and Anikoto are live and verified. Miruro is currently offline upstream — its pipe endpoint is down, and the debug routes above exist specifically to track when it comes back.