Soliplex
An AI-powered Retrieval-Augmented Generation (RAG) system with a modern web interface.
Overview
Soliplex is a self-hosted, full-stack RAG/AI system: a FastAPI backend (a thin configuration and authentication layer around pydantic-ai), a hardened Flutter client, and a terminal UI.
See the Overview for what Soliplex is, when (and when not) to use it, how it compares to alternatives, and the full feature and architecture breakdown.
Quickstart
The fastest way to stand up and run your own Soliplex is to let a skills-compatible AI agent (e.g. Claude Code, Codex, Copilot) do it for you. Soliplex publishes two Agent Skills that teach your agent how to deploy and operate it -- no need to clone or hack on this repository:
-
soliplex-templatescaffolds a complete, runnable Docker Compose stack (nginx, the Soliplex backend, the Flutter frontend, an ingester, Postgres, and a TUI). It prompts for the parameters you care about -- project name, host ports, your Ollama URL, models, version pins, auth mode -- and writes out the full project. Published from soliplex/soliplex-template. -
soliplex-docsis the full Soliplex documentation, so your agent can answer questions about installing, configuring, and operating your deployment straight from the official docs. Published from this repository.
Install the skills
Each skill is a tarball that unpacks into a soliplex-template/ or
soliplex-docs/ directory. Install them by unpacking both into whichever
directory your agent scans for skills. The directory differs per agent:
| Agent | Per user | Per project |
|---|---|---|
| Claude Code | ~/.claude/skills |
.claude/skills |
| OpenAI Codex | ~/.agents/skills |
.agents/skills |
| GitHub Copilot | ~/.copilot/skills |
.github/skills |
Point SKILLS_DIR at the row that matches your agent (the examples below use
the Claude Code per-user location) and unpack both skills into it:
SKILLS_DIR=~/.claude/skills
mkdir -p "$SKILLS_DIR"
# Deployment scaffolder
curl -fsSL \
https://github.com/soliplex/soliplex-template/releases/download/template-skill-latest/soliplex-template-skill.tar.gz \
| tar xz -C "$SKILLS_DIR"
# Documentation
curl -fsSL \
https://github.com/soliplex/soliplex/releases/download/docs-latest/soliplex-docs-skill.tar.gz \
| tar xz -C "$SKILLS_DIR"
For example, to install into a Codex project instead:
SKILLS_DIR=.agents/skills
The table covers the common cases; consult your agent's documentation (linked
above) for the authoritative list of locations it scans. Both skills bundle a
scripts/skill_versions.py helper to list published builds and upgrade in
place; see docs/docs-skill.md for details.
Use the skills
Start your agent in the directory where you want the deployment to live and
ask it to scaffold a stack -- for example, "Use the soliplex-template skill
to set up a new Soliplex deployment." The skill walks you through the
parameters and generates the project; then docker compose up brings it
online (the frontend defaults to http://localhost:9000).
Tell your agent to answer configuration or operations questions about Soliplex
using the soliplex-docs skill -- for example, "Use the soliplex-docs skill
to answer questions about configuring or operating a Soliplex installation."
and then "What is a Soliplex installation?"
The remainder of this README covers running Soliplex from a source checkout and developing on the backend itself. If you only want to use Soliplex, the two skills above are all you need.
Running from source
The skills-based Quickstart above is the recommended path for most users. To run Soliplex from a checkout of this repository instead, see:
- Prerequisites -- system requirements and an installation checklist.
- Server Setup -- install the backend, run the example installations, and start the server.
soliplex-clireference -- every subcommand and option (config audits, serving, admin users, model pulls, ...).- RAG Database -- create and populate the LanceDB database the examples expect.
- Client Setup -- run the Flutter web client.
- Docker Deployment -- extend the published container images.
Documentation
The full documentation set lives in docs/ and is published as a
Zensical site:
Tip: the same documentation is also available as the
soliplex-docsAgent Skill, so a skills-compatible agent can answer questions straight from it -- see the Quickstart above.
- Overview
- Prerequisites
- Server Setup and CLI Reference
- Client Setup
- Docker Deployment
- RAG Database
- Documentation Skill
- Configuration
- Usage
Configuration
Soliplex is configured by a tree of YAML files rooted at an
installation.yaml; sample configurations live in the
example/ directory. See the configuration docs:
- Filesystem Layout, Installation, Meta
- Rooms, Agents, Completions
- RAG, Quizzes, AI Skills, AG-UI Features
- Secrets, Environment, OIDC Authentication
- Logfire, Console logging, SQLAlchemy DBURIs
Development
Contributor guidance -- setup, build and test commands, code style, the configuration system, and how to add tools, rooms, and API endpoints -- lives in DEVELOPMENT.md. AGENTS.md and CLAUDE.md are condensed, agent-facing versions of the same guidance.
Related Repositories
- soliplex/frontend -- Flutter client (cross-platform mobile/desktop)
- soliplex/soliplex-template -- Docker Compose deployment template
- soliplex/ingester -- Content ingestion pipeline
- soliplex/ingester-agents -- Document ingestion agents
- soliplex/whitelabel -- Customer white-label appshell template
- Documentation site
License
MIT License - Copyright (c) 2025 Enfold Systems, Inc.
1 maintainer
Metadata
1 maintainer
Release files for soliplex 0.85.5
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| soliplex-0.85.5.tar.gz | 238.5 kB | Without provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| soliplex-0.85.5-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 516.8 kB
Release files / soliplex-0.85.5.tar.gz
| Download URL | soliplex-0.85.5.tar.gz |
|---|---|
| Size | 238.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
c8367b84916c20af2013851742ca339700e0b07250e06907acfc26acb73c4de8
|
|
BLAKE2b-256 checksum How to use checksums |
9dfb4aff0163da366f5e33e5b1ec837e44ce72752d3e420646d82bc2ba11fa95
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.13.1 {"installer":{"name":"uv","version":"0.13.1","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
Release files / soliplex-0.85.5-py3-none-any.whl
| Download URL | soliplex-0.85.5-py3-none-any.whl |
|---|---|
| Size | 278.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
242a303291212bb22dfd2372d00bd54661a9921f08e0b97f860990994ce7f5f4
|
|
BLAKE2b-256 checksum How to use checksums |
716c516b35f3407cb207be60901b1851d828187deb8d12472541610dd0bc297b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.13.1 {"installer":{"name":"uv","version":"0.13.1","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
Security for soliplex 0.85.5
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
