# Anime Akashic Records (AAR) API — Full Specification & Documentation

> High-precision semantic and hybrid discovery engine over 144,000+ canonical works spanning Anime, Manga, Light Novels, Visual Novels, Korean Manhwa, and Chinese Manhua. Designed to resolve vague memories, tropes, plot synopses, and aesthetic vibes into ranked canonical works paired with supporting excerpt passages.

---

## 1. Executive Summary & Recommended Positioning

### Recommended Positioning:
> **BEST PAIRED WITH TRADITIONAL ANIME APIS for exact metadata and images:**
> AAR is engineered to be paired seamlessly with traditional anime APIs (Jikan, AniList, Kitsu).
> Use traditional APIs for user watchlists, episode airing countdowns, and schedule trackers.
> Use AAR as the intelligence and semantic discovery layer for natural language trope search, vibe discovery, narrative recommendations, and AI agent integration.


> **Find the work you can't remember. Discover the work you didn't know you wanted.**
>
> Anime Akashic Records (AAR) is a discovery and metadata API for applications that need to search across works using natural-language descriptions, story concepts, and atmosphere — not only exact titles.

### Core Architecture & Key Strengths:
- **Six Unified Formats:** Cross-media mapping unifying Anime, Manga, Light Novels, Visual Novels, Korean Manhwa, and Chinese Manhua.
- **Hybrid Retrieval Pipeline:** Blends dense semantic vector retrieval with lexical keyword search and neural reranking for high-precision discovery.
- **Evidence-Grounded Results:** Every retrieved candidate returns supporting text from indexed synopses helping explain why a work matches (evaluated in context).
- **Question Answering (`/ask`):** Dedicated endpoint for conversational agents that synthesizes grounded prose explanations with direct citations.
- **High-Velocity Gateway:** Hosted on direct edge infrastructure (api.aarecords.tech) with instant API keys, sub-second responses, managed rate limits, and zero setup hassle.

### Intended Scope & Best Practices:
- **Core Purpose:** Semantic retrieval, vibe and trope discovery, natural language query resolution, and grounded AI answer synthesis across 144,000+ East Asian pop culture works.
- **Canonical Metadata Support:** Canonical metadata (`types`, `statuses`, release year, `genres`, `tags`, `studios`) is fully indexed, filterable, and returned in every payload to enable strict categorical scoping.
- **What AAR Is NOT:** AAR is not engineered as a real-time minute-by-minute episode airing tracker, schedule release countdown, or personal watch-list tracking manager. Use traditional encyclopedic APIs for live airing countdowns and watch-lists, and use AAR when you need intelligent discovery, semantic matching, and grounded reasoning.

---

## 2. Discovery Use Cases

Anime Akashic Records (AAR) is designed for more than identifying works by title. It supports natural-language discovery across anime, manga, light novels, visual novels, manhwa, and manhua.

### 2.1. Find a Work You Can't Remember
Describe a plot, scene, character, ability, or story detail without knowing the title. AAR searches for relevant works using semantic retrieval, keyword matching, and reranking.
- **Example:** `"An anime about a student who travels back in time to prevent a classmate's murder."`

### 2.2. Discover Works by Vibe
Describe the atmosphere, emotional tone, themes, or narrative experience you want to explore. You do not need to have seen the work before or remember a specific title.
- **Example:** `"Find fantasy stories with a quiet, melancholic atmosphere, slow journeys, and themes of memory and the passage of time."`

### 2.3. Search by Themes and Story Concepts
Find works involving specific narrative patterns, character dynamics, abilities, conflicts, or combinations of story elements.
- **Example:** `"Stories about an immortal protagonist who watches generations pass and struggles with the meaning of human relationships."`

### 2.4. Explore Connected Media
Discover related works across anime, manga, light novels, visual novels, manhwa, and manhua within AAR's indexed catalog. Availability and relationships depend on the indexed data.

### 2.5. Build AI-Powered Discovery Products
Developers can use AAR in AI chatbots, agents, community bots, recommendation interfaces, and discovery applications. Streaming platforms can use AAR's candidate results as an additional discovery layer and match them against their own available catalog.

### 2.6. Evidence-Grounded Results
AAR can return relevant work metadata and supporting text from indexed descriptions, helping applications explain why a result may match a query. Returned evidence should be evaluated in context; semantic relevance does not guarantee that every result is correct.

---

## 3. Commercial Target Audience & Product Positioning

Anime Akashic Records is **primarily an Anime API**, created to solve the discovery limitations of traditional databases.

### 3.1. The Problem with Traditional Anime APIs
Traditional anime and manga databases perform well for administrative catalogue management and user list logging, but exhibit fundamental failure modes when applied to modern discovery, recommendation systems, and AI agent grounding:

1. **Exact Title Lookup Focus:**
   Traditional APIs are engineered around exact or near-exact keyword matching. Searching for a half-remembered plot fragment like *"a boy whose arm transforms into a blade"* yields zero helpful results because the title is *Parasyte: The Maxim*.
2. **Rigid, Shallow Tag Checkboxes:**
   Filtering by broad genre categories like "Action" or "Fantasy" still leaves thousands of unrelated titles. Nuanced narrative themes like *"cozy winter solo travel"*, *"rivals who respect each other"*, or *"philosophical melancholy"* cannot be expressed with checkbox tags alone.
3. **Title-Only Output Without Context:**
   When a traditional system recommends an anime, it usually provides only a title string and an ID. Applications have no immediate way to explain *why* it fits the user query without reading external third-party reviews or scraping web forums.
4. **Cross-Medium Fragmentation:**
   Anime adaptations frequently originate from manga, light novels, visual novels, or webtoons. In conventional services, these works are fractured across disconnected databases, breaking cross-media continuity and adaptation lineage.

---

### 3.2. Why Choose Anime Akashic Records?
AAR is fundamentally distinct from conventional anime databases through six core architectural pillars:

1. **Natural Language Search:**
   Users and AI agents query in everyday language. Describe a character's powers, a memorable twist, or a specific aesthetic atmosphere, and the neural retrieval engine connects the description to relevant titles.
2. **Evidence-Grounded Results:**
   Results include supporting text passages from indexed synopses explaining *why* a work matches. This evidence enables users and AI assistants to evaluate recommendations with grounded context rather than blind guesswork.
3. **Mainly Anime — Plus Connected Mediums:**
   While focused primarily on Anime, the catalog unifies Japanese Manga, Light Novels, Visual Novels, Korean Manhwa, and Chinese Manhua in a single index. Multi-format franchises resolve cleanly across adaptations.
4. **Question Answering Endpoint (`/ask`):**
   A dedicated conversational endpoint tailored for AI chatbots and agents. Ask a direct question about an anime's premise or setup, and receive a factual, synthesized answer backed by catalog citations.
5. **Flexible Hard Scope Filters:**
   Combine the freedom of natural language descriptions with strict filters: narrow results by medium (`types`: anime, manga, light_novel, visual_novel, manhwa, manhua), status (Finished, Releasing), release year ranges, genres, or animation studios.
6. **Managed Search Infrastructure:**
   Hosted directly on production infrastructure (api.aarecords.tech). No need to host or maintain complex vector and retrieval clusters yourself — instant API access, managed rate limits, and ready-to-run client libraries.

---

### 3.3. At a Glance: How AAR Compares
Traditional APIs and Anime Akashic Records serve complementary roles in the anime ecosystem:

| Capability | Traditional Anime APIs (MAL / AniList / Kitsu) | Anime Akashic Records (AAR) |
|---|---|---|
| **Primary Focus** | User watch-lists, episode broadcast schedules, forum profiles. | Intelligent anime discovery, natural language search, vibe retrieval, and AI assistant grounding. |
| **Search Input** | Exact or near-exact title keywords only. | Plain natural language descriptions, plot summaries, tropes, vibes, and themes. |
| **Tip-of-the-Tongue & Vibe Queries** | Limited — requires title keywords to surface results. | Built specifically to resolve descriptive memories and atmospheric requests. |
| **Why It Matched (Evidence/Grounding)** | Returns titles and metadata IDs only. | Returns supporting synopsis excerpts providing context for why the title matches. |
| **Direct Q&A (`/ask`)** | Not available — you must build your own retrieval and grounding pipeline. | Built-in endpoint delivering synthesized answers with citations. |
| **Cross-Media Scope** | Often split into separate databases or IDs. | Unified index: Anime, Manga, Light Novels, Visual Novels, Korean Manhwa, Chinese Manhua. |
| **Best Used For** | Tracking which episode you watched yesterday. | Finding what to watch next, exploring vibes, answering questions, and powering intelligent apps. |

---

### 3.4. Target Personas & Integration Scenarios

| Target Audience | Practical Value Proposition | Technical Integration Note |
|---|---|---|
| **Streaming Platforms** | Add natural language search and vibe discovery to your existing catalog. Users describe what they want to watch in their own words. | Use AAR for candidate generation; match candidate titles against your own licensed catalog. *(Note: AAR surfaces candidate matches; presence in AAR does not imply streaming availability on any specific service).* |
| **AI Chatbots & Agents** | Equip your AI assistant with the ability to recommend anime accurately based on conversational user requests. | Synthesizes answers and returns cited excerpts to avoid hallucinations and unsupported claims. |
| **Anime Discovery Apps** | Enhance your app's search bar so users don't need to remember exact spellings; explore titles matching their mood or aesthetic vibe. | Replaces shallow tag checkboxes with rich conversational and thematic discovery. |
| **Community Bots & Creators** | Build Discord and Telegram bots that answer community questions, or quickly curate themed lists around specific tropes. | Rapidly surfaces thematic titles across 144,000+ indexed works. |

### 3.5. When to Choose AAR vs. Traditional APIs:
- **Choose Traditional APIs when:** You need user watch-list tracking (e.g. marking an episode watched), forum profiles, or minute-by-minute episode airing countdowns.
- **Choose AAR when:** You are building recommendation engines, AI chatbots, discovery search bars, or solving "what's that anime where..." and vibe-based discovery queries.

---

## 4. Gateway, Authentication & Base URLs

### Host & Endpoints
- **Production Gateway URL:** `https://api.aarecords.tech`
- **Public Free Key:** `aar_free_access` (or `aar_free`)
- **Interactive OpenAPI / Swagger UI:** `https://api.aarecords.tech/docs`
- **ReDoc Documentation:** `https://api.aarecords.tech/redoc`
- **Web Demo Playground:** https://meaha2000.github.io/aar-site/search.html
- **Official Telegram Bot:** https://t.me/AnimeAkashicRecords_bot (@AnimeAkashicRecords_bot)

### HTTP Authentication Headers
Requests to the production gateway support header, bearer token, or query parameter authentication:
```http
x-api-key: aar_free_access
Content-Type: application/json
Accept: application/json
```
*(Note: `GET /health` requires an API key, but does not consume your request quota).*

---

## 5. Endpoints Reference

### 5.1. `GET /health`
Liveness probe and catalog metrics. Requires an API key header (`x-api-key: aar_free_access`), but **does not consume request quota** (quota-free). Safe for container probes, uptime checkers, and integration smoke tests.

#### Request:
```bash
curl -X GET "https://api.aarecords.tech/health" \
  -H "x-api-key: aar_free_access"
```

#### Response (`200 OK`):
```json
{
  "status": "ok",
  "works": 144025,
  "chunks": 157340
}
```

---

### 5.2. `POST /search`
Advanced hybrid retrieval across the 6 supported mediums.

#### Request Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
| `question` | `string` | **Yes** | — | Natural language description, plot summary, character dynamic, or trope. |
| `limit` | `integer` | No | `8` | Number of ranked works to return (clamped between 1 and 50). |
| `mode` | `string` | No | `"hybrid"` | Retrieval mode: `"hybrid"` (semantic vectors + keyword search + rerank) or `"semantic"` (dense vector search + rerank). |
| `query_breakdown` | `boolean` | No | `true` | When `true`, decomposes and refines the query via LLM. When `false`, embeds directly as-is to minimize latency. |
| `nsfw` | `boolean` | No | `false` | Explicitly marks query as adult/NSFW, skipping LLM query refinement and proceeding directly to vector embedding. |
| `include` | `array[string]` | No | `[]` | Extra relational fields to include: `["staff", "characters", "relations"]`. |
| `scope` | `object` | No | `{}` | Hard filters applied to the index before retrieval. |

#### Scope Filter Object Schema (`scope`)
| Scope Field | Type | Description & Allowed Values |
|---|---|---|
| `types` | `array[string]` | Filter mediums: `["ANIME", "MANGA", "LIGHT_NOVEL", "VISUAL_NOVEL", "MANHWA", "MANHUA"]`. |
| `statuses` | `array[string]` | Lifecycle status: `["finished", "releasing", "upcoming", "cancelled", "hiatus"]` (also accepts `"completed"`, `"airing"`, `"ongoing"`). |
| `genres` | `array[string]` | Genres: `["Action", "Adventure", "Comedy", "Drama", "Fantasy", "Horror", "Mecha", "Mystery", "Psychological", "Romance", "Sci-Fi", "Slice of Life", "Sports", "Supernatural", "Thriller"]`. |
| `tags` | `array[string]` | Thematic tags: `["Time Travel", "Cyberpunk", "Isekai", "Reincarnation", "Cultivation", "Martial Arts", "Dungeons", "Necromancy", "School", "Anti-Hero"]`. |
| `themes` | `array[string]` | Narrative themes and tropes. |
| `studios` | `array[string]` | Studio name or substring: `["ufotable", "MAPPA", "Kyoto Animation", "Shaft", "Madhouse", "Bones", "Wit Studio"]`. |
| `director_names` | `array[string]` | Credited directors (e.g. `["Shinichirou Watanabe", "Hideaki Anno"]`). |
| `author_names` | `array[string]` | Original creators, mangaka, or novelists (e.g. `["Nisio Isin", "Naoki Urasawa"]`). |
| `composer_names` | `array[string]` | Music composers (e.g. `["Yoko Kanno", "Hiroyuki Sawano"]`). |
| `voice_actor_names` | `array[string]` | Voice actors / seiyuu (e.g. `["Megumi Hayashibara", "Hiroshi Kamiya"]`). |
| `staff_names` | `array[string]` | Any credited production staff member. |
| `character_names` | `array[string]` | Canonical character names (e.g. `["Spike Spiegel", "Rintaro Okabe"]`). |
| `titles` | `array[string]` | Restrict search to specific work titles. |
| `years` | `array[integer]` | Specific release years (e.g. `[1998, 2011]`). |
| `min_year` | `integer` | Minimum release year bound (e.g. `2010`). |
| `max_year` | `integer` | Maximum release year bound (e.g. `2024`). |
| `min_score` | `number` | Minimum review score threshold (float between 0.0 and 10.0). |
| `adult` | `boolean` | Explicit 18+ content flag (default `false` excludes adult titles). |

#### Request Example:
```json
{
  "question": "A hunter who can see hidden status windows and level up alone in modern dungeons",
  "limit": 8,
  "mode": "hybrid",
  "query_breakdown": true,
  "scope": {
    "types": ["MANHWA", "MANGA", "ANIME"],
    "statuses": ["finished", "releasing"],
    "genres": ["Action", "Fantasy"]
  }
}
```

#### Response Envelope (`200 OK`):
```json
{
  "total_documents": 1,
  "results": [
    {
      "work_id": "manhwa:solo_leveling",
      "title": "Solo Leveling",
      "type": "manhwa",
      "score": 0.942,
      "status": "finished",
      "chunks": [
        {
          "section": "Synopsis",
          "text": "In a world where hunters must battle deadly monsters, Sung Jin-woo, notoriously known as the weakest hunter of all mankind, is fatally injured in a high-rank double dungeon..."
        }
      ]
    }
  ],
  "trace": {
    "total_time_ms": 284.5
  }
}
```

---

### 5.3. `POST /ask`
Grounded conversational question-answering with citation synthesis for AI agents.

#### Request Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
| `question` | `string` | **Yes** | — | Question regarding an anime premise, comparison, or plot inquiry. |
| `limit` | `integer` | No | `8` | Number of ranked works to return (clamped between 1 and 50). |
| `mode` | `string` | No | `"hybrid"` | Retrieval mode: `"hybrid"` or `"semantic"`. |
| `query_breakdown` | `boolean` | No | `true` | When `true`, decomposes query via LLM. When `false`, embeds query directly as-is to minimize latency. |
| `nsfw` | `boolean` | No | `false` | Explicitly marks query as adult/NSFW, skipping LLM query refinement. |
| `include` | `array[string]` | No | `[]` | Extra relational fields to attach: `["staff", "characters", "relations"]`. |
| `scope` | `object` | No | `{}` | Optional hard scope filters restricting candidate retrieval. |

#### Request Example:
```json
{
  "question": "What is the premise of Parasyte and what makes its relationship dynamic unique?"
}
```

#### Response Envelope (`200 OK`):
```json
{
  "answer": "In Parasyte (Kiseijuu: Sei no Kakuritsu), seventeen-year-old Shinichi Izumi is partially infected by an alien parasite named Migi, which fails to take over his brain and instead matures in his right hand. The dynamic is unique because Shinichi and Migi retain separate intellects and moralities, forcing an uneasy symbiotic coexistence where Migi's cold survival logic clashes with Shinichi's human empathy as they defend against other hostile parasites.",
  "sources": [
    {
      "ref": 1,
      "title": "Parasyte -the maxim-",
      "type": "anime",
      "section": "Synopsis",
      "confirmed": true
    }
  ],
  "results": [
    {
      "work_id": "anime:parasyte_the_maxim",
      "title": "Parasyte -the maxim-",
      "type": "anime",
      "score": 0.951,
      "status": "finished",
      "chunks": [
        {
          "section": "Synopsis",
          "text": "Seventeen-year-old Shinichi Izumi falls victim to one of these parasites, but it fails to take over his brain, ending up in his right hand instead."
        }
      ]
    }
  ]
}
```

---

---

### 5.4. `POST /recommend` & `GET /recommend/{work_id}`
Work-to-work semantic recommendation based on narrative DNA, thematic essence, character dynamics, and emotional tone.

#### Request Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
| `seed_work_ids` | `array[string]` | **Yes** (for POST) | — | 1 to 10 canonical work IDs (e.g. `["work:anime:mal_9253_al_9253"]`). |
| `work_id` | `string` | **Yes** (for GET) | — | Path parameter in `/recommend/{work_id}`. |
| `limit` | `integer` | No | `10` | Number of recommendations to return (1–50). |

#### Request Example (GET):
```bash
curl -X GET "https://api.aarecords.tech/recommend/work:anime:mal_9253_al_9253?limit=5" \
  -H "x-api-key: aar_free_access"
```

#### Request Example (POST):
```json
{
  "seed_work_ids": ["work:anime:mal_9253_al_9253"],
  "limit": 5
}
```

#### Response Envelope (`200 OK`):
```json
{
  "work_id": "work:anime:mal_9253_al_9253",
  "title": "Steins;Gate",
  "total_recommendations": 5,
  "recommendations": [
    {
      "position": 1,
      "work_id": "work:anime:mal_30484_al_21127",
      "title": "Steins;Gate 0",
      "type": "anime",
      "similarity_score": 0.9412
    },
    {
      "position": 2,
      "work_id": "work:anime:mal_22535_al_20665",
      "title": "Kiseijuu: Sei no Kakuritsu",
      "type": "anime",
      "similarity_score": 0.8125
    }
  ]
}
```

---

### 5.5. `POST /suggest` & `GET /contributions/grouped`
Community proposals and catalog contributions endpoint. Allows users and AI agents to submit missing titles, rich narrative descriptions, or metadata corrections.

> **🤝 We Actively Welcome Contributions Here:**
> We actively welcome community and AI agent contributions through this endpoint — whether proposing missing titles, providing rich narrative plot descriptions, or suggesting corrections to existing works. All submissions are automatically clustered, deduplicated, and reviewed for inclusion in canonical catalog updates.

#### Request Parameters (`POST /suggest` or `POST /contributions`):
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
| `contribution_type` | `string` | **Yes** | — | One of `"ADD_WORK"`, `"ADD_ALIAS"`, `"EDIT_METADATA"`, `"FEATURE_OR_VIBE"`. |
| `media_type` | `string` | **Yes** | — | One of `"anime"`, `"manga"`, `"light_novel"`, `"visual_novel"`, `"manhwa"`, `"manhua"`. |
| `title` | `string` | **Yes** | — | Canonical or primary title. |
| `custom_description` | `string` | No | `""` | Detailed narrative synopsis, worldbuilding, character dynamics (up to 10k chars). |
| `aliases` | `array[string]` | No | `[]` | Alternative names, Romaji titles, or English titles. |
| `author` | `string` | No | `""` | Original creator, author, or director. |
| `release_year` | `integer` | No | `null` | Release year (e.g. `2024`). |
| `tags` | `array[string]` | No | `[]` | Thematic tags or genres. |

#### Request Example (`POST /suggest`):
```bash
curl -X POST "https://api.aarecords.tech/suggest" \
  -H "Content-Type: application/json" \
  -H "x-api-key: aar_free_access" \
  -d '{
    "contribution_type": "ADD_WORK",
    "media_type": "anime",
    "title": "Sousou no Frieren",
    "custom_description": "An elf mage embarks on a peaceful, decades-long journey reflecting on human mortality and connection.",
    "alt_titles": ["Frieren: Beyond Journey\'s End"],
    "author": "Kanehito Yamada",
    "release_year": 2023,
    "suggested_tags": ["Fantasy", "Adventure", "Drama"]
  }'
```

#### Inspect Clustered Submissions (`GET /contributions/grouped`):
```bash
curl -X GET "https://api.aarecords.tech/contributions/grouped" \
  -H "x-api-key: aar_free_access"
```

## 6. Client Code Snippets

### Python (`requests`)
```python
import os
import requests

url = "https://api.aarecords.tech/search"
headers = {
    "x-api-key": os.getenv("AAR_API_KEY", "aar_free_access"),
    "Content-Type": "application/json"
}
payload = {
    "question": "A girl traveling through desolate ruins with two small robotic companions",
    "limit": 5,
    "scope": {
        "types": ["ANIME", "MANGA"],
        "genres": ["Sci-Fi", "Adventure"]
    }
}

response = requests.post(url, json=payload, headers=headers, timeout=10)
response.raise_for_status()
data = response.json()

for idx, item in enumerate(data.get("results", []), start=1):
    print(f"#{idx} {item['title']} ({item['type']})")
    for chunk in item.get("chunks", []):
        print(f"   [{chunk.get('section')}] {chunk.get('text', '')[:120]}...")
```

### TypeScript / JavaScript (Node.js Fetch)
```typescript
interface SearchPayload {
  question: string;
  limit?: number;
  scope?: {
    types?: string[];
    statuses?: string[];
    genres?: string[];
  };
}

async function searchAAR(payload: SearchPayload) {
  const response = await fetch("https://api.aarecords.tech/search", {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      "x-api-key": process.env.AAR_API_KEY || "aar_free_access"
    },
    body: JSON.stringify(payload)
  });

  if (!response.ok) {
    throw new Error(`API error: ${response.status} ${response.statusText}`);
  }

  return await response.json();
}
```

### cURL CLI
```bash
curl -X POST "https://api.aarecords.tech/search" \
  -H "Content-Type: application/json" \
  -H "x-api-key: aar_free_access" \
  -d '{
    "question": "Time-travel visual novel with branching timelines",
    "limit": 5,
    "scope": { "types": ["VISUAL_NOVEL"] }
  }'
```

---

## 7. Status Codes & Error Envelopes

- **`200 OK`**: Request succeeded. Returns the JSON envelope.
- **`401 Unauthorized`**: Key rejected or missing header:
  ```json
  {"detail": "missing or invalid API key"}
  ```
- **`422 Unprocessable Entity`**: Payload validation failed (e.g., unknown scope filter, empty question string).
- **`429 Too Many Requests`**: Rate limit exceeded for your tier (30 req/min for `/search`, 3 req/min for `/ask`). Reset occurs per rolling window.

---

## 8. Official Telegram Bot

A full-featured live client is available on Telegram for conversational testing and interactive lookups:
- **Bot Username:** `@AnimeAkashicRecords_bot`
- **Link:** https://t.me/AnimeAkashicRecords_bot
- **Commands:**
  - `/s <query>` or `/search <query>`: Ranked search across all 6 media formats.
  - `/a <question>` or `/ask <question>`: Grounded AI question answering.
  - `/scope <filters>`: Sets conversational filter parameters (e.g. `/scope type=manhwa,genre=Action`).
  - `/work <title>`: Direct canonical record and chunk inspector.
  - `/stats`: Live corpus statistics and index status.

---

## 9. Developer Support, Collaborations & Feedback

We welcome questions, developer integrations, collaboration ideas, bug reports, and feature proposals:
- **Official Contact Email:** `AnimeAkashicRecords@hotmail.com`
- **GitHub Repository:** https://github.com/Meaha2000/aar-site
- **OpenAPI / Swagger UI:** https://api.aarecords.tech/docs