Skip to content

nneibaue/shadow-os

Repository files navigation

Shadow OS

This package contains comprehensive planning documents for Wendell's "mech suit" cognitive amplification system. Everything here is textual (Markdown) with detailed pseudocode — no implementation code, but ready for development.

Contents

  • 01_nate_understanding.md — Enhanced understanding including the deeper context of shadow work, coaching methodology, and the "grammar of ontology" vision.
  • 02_questions_for_wendell.md — Expanded questions covering multi-modal capture, relationship dynamics, and configuration approaches.
  • 03_design_plan.md — Comprehensive design with enhanced config, sophisticated data models, and detailed operational flows.
  • 04_research_summary.md — RAG and agentic techniques specifically tailored for speed-of-thought cognitive amplification.
  • 05_prompts_catalog.md — Production-ready system prompts with detailed instructions for each agent, including new voice transcription and book generation capabilities.
  • 06_agents_pseudocode_and_prompts.md — Complete agent specifications with enhanced tools, sophisticated pseudocode, and conversation-informed system prompts.

Key Architectural Insights from Conversation

  • Pydantic AI as the mandated agent framework
  • YAML configuration so Wendell can "tweak the knobs" without code changes
  • VS Code + Copilot/MCP integration for development acceleration
  • Contradiction embrace as a feature, not a bug
  • Citation integrity as intellectual honesty enforcement

The "Mech Suit" Vision

This system is designed to be "the mech suit that Wendell needs to do [information extraction and composition] as efficiently and as effectively as possible." It bridges:

  • Immediate insight capture (the "that's a Bar" moment)
  • Structured knowledge building (relationships, expansions, evidence)
  • On-demand book generation (AI agents that can "produce a book at any moment's notice")

Implementation Readiness

All prompts aim to be production-ready for Claude Sonnet. All pseudocode is detailed enough for direct translation to Pydantic AI agents. The configuration schemas are complete and ready for YAML implementation.

Note: Bars are immutable atoms with memetic quality; expansions and relations are editable, versioned artifacts that preserve intellectual honesty through rigorous citation.tarter Package (v2025-08-26)

This package contains planning documents for Wendell’s authoring system. Everything here is textual (Markdown) with pseudocode only — no real code.

Note: Bars are immutable atoms; expansions and relations are editable, versioned artifacts. Config knobs are front-facing (YAML), so Wendell can tweak behavior without touching code.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors