Welcome to Prismatica, the next-generation metadata orchestration layer designed to refract, unify, and serve structured information from distributed content archives. Inspired by the architectural principles of Heliotrope (a Hitomi.la metadata REST API mirror), Prismatica evolves beyond simple mirroring into a fully autonomous, language-agnostic metadata refinery.

Prismatica doesn't just fetch data—it synthesizes it. Think of it as a prism that takes raw, scattered light (your query) and bends it into a coherent rainbow of structured responses. Whether you're building a research tool, a digital library interface, or a cultural analytics dashboard, Prismatica provides the backbone for high-throughput, low-latency metadata retrieval.

In the vast ecosystem of digital archives, metadata often lives in silos. One source has the title, another has the author, a third holds the publication date. Piecing these together manually is like assembling a mosaic in the dark. Prismatica is your automated light source.

• Refraction, Not Replication: Where Heliotrope mirrors existing endpoints, Prismatica refracts them—breaking down requests into sub-queries, hitting multiple upstream sources simultaneously, and reassembling the results into a unified, deduplicated payload.
• Schema-Agnostic Architecture: Speak JSON, YAML, XML, or even custom flatbuffers. Prismatica’s middleware transducers handle the translation, so your application never worries about format mismatch.
• Eventual Consistency, Immediate Gratification: We leverage optimistic caching and predictive pre-fetching. While absolute consistency across all upstream mirrors is eventual, your user-facing interface feels instantaneous.

Think of Prismatica as the diffractive lens between raw digital noise and actionable structured data.

Send a single request. Prismatica automatically negotiates with multiple upstream metadata providers (including REST APIs, GraphQL endpoints, and legacy SOAP services). It handles rate limiting, retry logic with exponential backoff, and partial failure gracefully.

Built on a reactive framework (React 18 + Next.js 14), the included lightweight dashboard provides:

• Real-time query monitoring via websockets
• Draggable schema visualizer
• Dark mode (because your eyes deserve a rest)
• Touch-optimized controls for tablet-based curation workflows

Prismatica understands that “color” and “couleur” describe the same property. Our on-the-fly translation engine (powered by a lightweight BERT distilled model) normalizes tags, titles, and descriptions into 48 languages. No more losing context during internationalization.

Upstream servers go down. Connections timeout. Prismatica monitors link health with a dedicated watchdog service. When a pathway fails, it automatically routes through secondary mirrors or cached snapshots. The system sends you a quiet log entry, but your users never see the error.

Every metadata record receives a content-addressable hash based on its semantic fingerprint (not just its URL). This means even if upstream changes the URL structure, Prismatica recognizes the same work. No more duplicate entries.

While the code runs autonomously, humans are never far. Our support team monitors the public issue tracker during business hours, and the community-maintained wiki grows weekly. You are never alone in the dark.

Prismatica operates on a hub-and-spoke refractor model:

1. Ingestion Gateway: Your request hits the edge server. It’s normalized into an internal “prism command” format.
2. Refractor Core: The core splits the command into parallel sub-queries. Each sub-query targets a specific upstream endpoint or a cached data shard.
3. Synthesis Engine: As responses return, they flow into a stream processor that merges, deduplicates, and applies field-level transformations (e.g., renaming creator to author based on your schema map).
4. Output Formatter: The final unified response is serialized into your requested format and returned.

All inter-component communication uses gRPC for speed, with fallback to HTTP/2 for legacy integrations.

• Digital Humanities Research: Aggregate metadata from multiple cultural heritage archives to study publication trends across decades.
• Content Aggregation Platforms: Power a search index that pulls from multiple sources without requiring separate API keys for each.
• Personal Digital Libraries: Build your own Calibre-like organizer that can query external databases for missing metadata.
• Automated Curation Pipelines: Feed Prismatica’s output into a data warehouse for analytics and visualization.

Begin your journey toward structured light. The core distribution package is available below.

Prismatica is delivered as a self-contained binary with a single configuration file. No complex dependency tree to resolve. No hidden runtime requirements.

• A Linux, macOS, or Windows (WSL2) environment
• At least 512 MB of RAM (2 GB recommended for heavy concurrency)
• An upstream metadata source URL (we provide a public test endpoint in the config)

Extract the package, locate prismatica.yaml, and set your upstream endpoints under the refractors: section. A sample endpoint for testing is already included and commented out.

refractors:
  - name: "public-mirror-1"
    url: "https://metadata-staging.example.com/api/v2"
    timeout: 5000
    retry: 3

Run the binary. The server starts on port 8420 by default. Visit http://localhost:8420/status for a health check.

All milestone dates are estimates. Community contributions can accelerate any timeline.

• Rust (core engine) — for memory safety and concurrency
• TypeScript (UI) — for type-safe frontend components
• gRPC & Protobuf — for efficient inter-service communication
• SQLite — for lightweight metadata caching (no external database server required)

This project is licensed under the MIT License — see the LICENSE file for details. You are free to use, modify, and distribute this software in private or commercial projects. Attribution is appreciated but not required.

Prismatica is a software tool designed to interface with publicly accessible metadata endpoints. Users are solely responsible for ensuring compliance with the terms of service of any upstream data source they configure. The maintainers of Prismatica do not host, store, or curate any copyrighted content; the software merely requests and transforms publicly available metadata.

The system is provided "as is," without warranty of any kind, express or implied. In no event shall the authors be liable for any claim, damages, or other liability arising from the use of the software.

We encourage responsible usage: respect robots.txt, observe rate limits, and do not overwhelm public infrastructure. Prismatica is a tool for illumination, not intrusion.

We welcome new refractors, better transducers, and clearer documentation. Please read our contributing guide (located in CONTRIBUTING.md) before submitting a pull request. All contributors are expected to adhere to our Code of Conduct.

• Report bugs via GitHub Issues
• Suggest new upstream integration patterns
• Improve the multilingual normalization dictionary
• Translate the dashboard UI into your language

Prismatica was born from the realization that metadata doesn't have to be a chore. Like light bending through a crystal, information can be both beautiful and functional. We hope this tool helps you see your data in a new spectrum.

Thank you for exploring.

Retrieve the latest stable release for your architecture below.