Skip to content
View AdrianParedez's full-sized avatar

Block or report AdrianParedez

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.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. 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
AdrianParedez/README.md

Adrian Paredez

First-principles systems · hardened runtimes · open-source engineering

paredez.dev · milner · capability-fabric

Rust Windows Agent Skills Dependencies Public work

I build software from first principles: small surfaces, explicit contracts, minimal dependencies, and behaviour that can be inspected under pressure. My current public work centres on Milner, a Windows-first command runtime, and Capability Fabric, a model-agnostic Agent Skills library for explicit routing, bounded context, and verifiable agent behaviour.

Public Focus

Milner A Windows-first Rust command runtime built around explicit process launch, typed parsing, narrow execution policy, and hardened operating-system boundaries.

Repository · Latest release
Capability Fabric A model-agnostic Agent Skills library for routing, planning, research, context budgeting, sustained execution, verification, and skill authoring.

Repository · Latest release

Install Milner on Windows:

irm https://paredez.dev/install.ps1 | iex

Current Lane

native execution      typed command surfaces      first-principles systems
hardened runtimes     agent operating method      safety by construction
minimal dependencies   open-source engineering     verifiable behaviour
Working principles
  • Design from first principles before importing complexity.
  • Keep dependencies intentional, minimal, and justified.
  • Prefer explicit contracts over ambient behaviour.
  • Reject ambiguous input instead of guessing.
  • Make verification part of the operating surface.
  • Keep agent work routed, bounded, and inspectable.
  • Optimise from evidence, not instinct.
  • Publish only what can be maintained with discipline.
Recurring stack
  • Rust for command runtimes and native execution boundaries
  • TypeScript for web-facing control surfaces
  • Tauri for planned desktop applications
  • Python for research automation and local AI experiments
  • Cloudflare for lightweight public infrastructure
  • Agent Skills for portable agent operating methods

Built with assistance from Sole.

Popular repositories Loading

  1. AdrianParedez AdrianParedez Public

    Public profile README for Adrian Paredez

  2. milner milner Public

    A Windows-first Rust shell experiment built around explicit process launch, typed parsing, and narrow execution policy.

    Rust

  3. capability-fabric capability-fabric Public

    A model-agnostic Agent Skills library for explicit routing, bounded context, and verifiable agent behaviour.

    PowerShell

  4. rootline rootline Public

    A route-aware AGENTS.md protocol for AI agents.