Flux.Abstractions
0.24.0
dotnet add package Flux.Abstractions --version 0.24.0
NuGet\Install-Package Flux.Abstractions -Version 0.24.0
<PackageReference Include="Flux.Abstractions" Version="0.24.0" />
<PackageVersion Include="Flux.Abstractions" Version="0.24.0" />
<PackageReference Include="Flux.Abstractions" />
paket add Flux.Abstractions --version 0.24.0
#r "nuget: Flux.Abstractions, 0.24.0"
#:package Flux.Abstractions@0.24.0
#addin nuget:?package=Flux.Abstractions&version=0.24.0
#tool nuget:?package=Flux.Abstractions&version=0.24.0
Flux.Abstractions
Shared contracts for the Flux RAG ecosystem. Pure interfaces, zero dependencies.
dotnet add package Flux.Abstractions
What is in here
| Type | Purpose |
|---|---|
IEnrichedChunk |
A document chunk plus the metadata a retrieval layer needs to rank and cite it. |
ISourceMetadata |
Where a chunk came from, for traceability and filtering. |
ILanguageProfile |
The minimal language-specific segmentation metadata every module agrees on. |
ITextCompletionService |
A text-completion port, so a pipeline stage can call a model without binding to a provider. |
TextCompletionOptions |
Sampling options for the above. |
IEnrichedChunk and ISourceMetadata are deliberately a superset union of what the
individual modules produce: an implementation returns null or the default for fields that do
not apply to it. That keeps one contract instead of one per producer, at the cost of some
properties being empty depending on the source.
ILanguageProfile is intentionally minimal. Modules that need more extend it privately rather
than widening the shared contract — a producing module's richer profile stays in that module.
Why this package has no dependencies
Every package in the ecosystem sits above this one. Anything referenced here becomes a version
floor for all of them, so the absence of dependencies is a constraint the package is designed
around, not an accident of it being small. Contributions that add a PackageReference here
should expect that question first.
The same reasoning explains why this is its own repository. A contract package must be able to sit below its consumers in the dependency graph. While it shipped from inside one of the packages that consumes it, that was not possible: the consumers could only ever reference a release older than the one they were part of, and the graph had a cycle in it. Splitting the contract out removes the cycle and lets consumers track the current contract.
Versioning
This package continues the version line it published under previously — the package id is unchanged and the sequence keeps moving forward, so upgrading is a normal version bump with no migration step. The independent line starts at 0.24.0.
Being a contract package, it changes rarely, and it now versions on its own cadence instead of inheriting a release number from a package that ships far more often.
Pre-1.0: breaking changes are possible on minor bumps, and are called out in the release notes.
License
MIT
| Product | Versions Compatible and additional computed target framework versions. |
|---|---|
| .NET | net10.0 is compatible. net10.0-android was computed. net10.0-browser was computed. net10.0-ios was computed. net10.0-maccatalyst was computed. net10.0-macos was computed. net10.0-tvos was computed. net10.0-windows was computed. |
-
net10.0
- No dependencies.
NuGet packages (6)
Showing the top 5 NuGet packages that depend on Flux.Abstractions:
| Package | Downloads |
|---|---|
|
FileFlux
Complete document processing SDK optimized for RAG systems. Transform PDF, DOCX, Excel, PowerPoint, Markdown and other formats into high-quality chunks with intelligent semantic boundary detection. Includes advanced chunking strategies, metadata extraction, and performance optimization. |
|
|
FluxIndex.Core
FluxIndex Core - Domain models and application interfaces for FluxIndex RAG system |
|
|
FileFlux.Core
Pure document extraction SDK for RAG systems. Zero AI dependencies. Extract text from PDF, DOCX, Excel, PowerPoint, Markdown, HTML, and text files. Provides IDocumentReader interface and implementations. Use FileFlux.Core for extraction-only scenarios. For AI-enhanced extraction (image OCR, captioning), use the FileFlux package. |
|
|
FluxCurator.Core
Core library for FluxCurator - Minimal dependencies. PII masking, content filtering, and rule-based chunking with Korean language support. |
|
|
FluxImprover
The Quality Layer for RAG Data Pipelines - LLM-powered enrichment and quality assessment |
GitHub repositories
This package is not used by any popular GitHub repositories.
| Version | Downloads | Last Updated |
|---|---|---|
| 0.24.0 | 1,223 | 8/2/2026 |
| 0.23.0 | 268 | 7/30/2026 |
| 0.21.5 | 239 | 7/28/2026 |
| 0.21.4 | 224 | 7/28/2026 |
| 0.21.3 | 239 | 7/28/2026 |
| 0.21.2 | 225 | 7/28/2026 |
| 0.21.1 | 212 | 7/28/2026 |
| 0.21.0 | 508 | 7/28/2026 |
| 0.20.2 | 217 | 7/24/2026 |
| 0.20.1 | 196 | 7/22/2026 |
| 0.20.0 | 206 | 7/21/2026 |
| 0.19.0 | 208 | 7/18/2026 |
| 0.18.0 | 191 | 7/17/2026 |
| 0.17.1 | 195 | 7/17/2026 |
| 0.17.0 | 412 | 7/3/2026 |
| 0.16.0 | 219 | 7/2/2026 |
| 0.15.0 | 243 | 6/28/2026 |
| 0.14.0 | 219 | 6/19/2026 |
| 0.13.21 | 224 | 6/15/2026 |
| 0.13.20 | 224 | 6/15/2026 |