Outrage
A simple RAG (Retrieval-Augmented Generation) system for LLMs, designed to just work with minimal setup, and provide a memory that interoperates between coding harnesses such as Claude Code, Codex, Cursor, and Copilot CLI.
Features
- Simple to get started with - built in initialisation configures everything.
- Hierarchical key based store of data and arbitrary metadata.
- Search by exact text, by metadata, or optionally by meaning with a small classification model that screens title, keywords, summary, contents and body in turn. Search by meaning is experimental.
- Multiple data stores including parquet based stores for large-scale reference material, and an experimental PostgreSQL store shared by several devices.
- Explicit support for command-line coding agents: Claude Code, OpenAI Codex CLI, Cursor Agent, and GitHub Copilot CLI.
- Command line tools for manipulating the data stores.
Getting started
Set up a 3.12 or later Python environment, then:
pip install "outrage[all]"
To set up a project to use Outrage, change directory to the root of the project and run:
outrage init
And that's it. That will set up the MCP server, tools, skills, agents and hooks such that when you run your LLM harness the agent will use the Outrage store as a memory. Getting started has more.
Guides
- Getting started
- Setting up projects
- Mounting data stores
- Importing documents
- Export and backup
- Parquet stores
- Postgres - experimental.
- Smart search - experimental.
Documentation
All of it is in src/outrage/documents, a documentation tree written to be read
through the store: it ships in the wheel, and the MCP server mounts it
read-only at the outrage key, so a session reads the manual with the same
tools it reads everything else with. It is ordinary markdown as well, so the
links here work too.
- the documents - the top of that tree, and where to start. The guides, keys, the command line, the tools and the conventions all hang off it.
- the command line and the MCP tools - the reference for each front end.
- the design - why Outrage is built the way it is.
- development - the development environment, for working on Outrage itself.
1 maintainer
Metadata
1 maintainer
Release files for outrage 1.3.1
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 | |
|---|---|---|---|---|
| outrage-1.3.1.tar.gz | 1.5 MB | Without provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| outrage-1.3.1-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 2.5 MB
Release files / outrage-1.3.1.tar.gz
| Download URL | outrage-1.3.1.tar.gz |
|---|---|
| Size | 1.5 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
72b5f48b7b2ef0a7847c2620b6fe2d636b220e567b2e7ea4a60b357c444a94c4
|
|
BLAKE2b-256 checksum How to use checksums |
cf6d2ba74f18a8aeecec268968ac95b6b7b0b0a645cfce6e586b502446779de8
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.10.14
|
Release files / outrage-1.3.1-py3-none-any.whl
| Download URL | outrage-1.3.1-py3-none-any.whl |
|---|---|
| Size | 1.0 MB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
7d236020e57ea1d157cb6a2bdebbb27390ec3dc706fb9cf0f1526f4bbf4ad269
|
|
BLAKE2b-256 checksum How to use checksums |
e1c8b582f2243bda289b4aed9c7e23d1b5996c561144ffbcc23fc720eefb233b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.10.14
|
Security for outrage 1.3.1
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.
