Skip to content
View LuminariSoftwares's full-sized avatar

Block or report LuminariSoftwares

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
LuminariSoftwares/README.md

Luminari Softwares -- four small MIT-licensed tools for people running local models. Each does one job, shares no code with the others, and works alone. All four are in active development.

  • tool-guardian -- sits in front of all your MCP servers and loads each tool's schema only when it's needed.
  • context-guardian -- compacts long local-model sessions before the window fills, archives the originals, fails open.
  • stack-guardian -- a health doctor for a home AI stack, with findings that earn confidence by repetition.
  • code-guardian -- scores a repository Safe / Works / Complete before you run it, without executing it.

Python standard library where possible, hand-written MCP servers, CI on Linux, Windows and macOS, a privacy scan on every push. One maintainer; issues welcome, acknowledged within a day.

Pinned Loading

  1. tool-guardian tool-guardian Public

    Too many MCP servers eating your context? tool-guardian sits in front of all of them and loads each tool’s schema only when needed. A 7-server setup went from 20,005 tokens to 1,084. Any MCP client.

    Python 2

  2. context-guardian context-guardian Public

    40 turns: 1,130,722 tokens upstream -> 161,141. A fail-open compaction proxy for local OpenAI-compatible models.

    Python 2

  3. code-guardian code-guardian Public

    Safe to run? Will it work? Complete? An A-F grade for any repo, without running it. 0 false FAILs on 8 well-known repos. CLI + MCP.

    Python

  4. stack-guardian stack-guardian Public

    14 checks, one command, graded findings. A health doctor for a home AI stack: Ollama, NVIDIA GPUs, tailnet.

    Python