These guides document the GSD (Get Shit Done) workflow system and the Skill Creator adaptive learning layer. The architecture draws from Amiga chipset design — specialized coprocessors coordinating through a shared bus, each handling a different domain of the computation.
Guides & Documentation
Getting Started Guide
First steps with GSD and Skill Creator. Installation, configuration, your first project, understanding the workflow. Start here if you're new.
Advanced Guide
Deep configuration, custom chipsets, multi-agent orchestration, convoy operations, performance tuning, and production deployment patterns. For experienced users scaling up.
SC:Learn Pipeline Guide
How SC:Learn transforms external knowledge sources into reusable skills for Claude Code. Seven-stage pipeline: Acquire → Sanitize → HITL Gate → Analyze → Extract → Dedup → Generate. Three-layer safety architecture. Walkthrough with Blender docs example.
The Chipset — Sysadmin Guide
How the Skill Creator chipset maps hardware architecture to AI agent orchestration. 14 sections covering: Amiga blueprint, context window as system RAM, budget as memory management, bus architecture, DMA knowledge pipeline, Exec kernel scheduling, Copper coprocessor, and tuning diagnostics.
Chipset Circuit Diagram
Visual SVG schematic of the Gastown coprocessor architecture. Four chips: Agnus (context/DMA, 60%), Denise (rendering, 15%), Paula (I/O, 15%), Gary (routing, 10%). Central Claude CPU/Exec Kernel. Copper Pipeline, Blitter Offload, 14 agents, 5 teams, 11-event protocol.
Git, GSD & Skill Creator — Project Management
Complete git workflow guide integrated with GSD. 16 sections: version control fundamentals, branching, merging, conflict resolution, worktrees, deployment pipelines, FHS folder organization, document intake, team collaboration, agent isolation, and command recipes.
GSD User Guides Suite
Community documentation for GSD v1.35.0, released under CC BY-SA 4.0 by the Tibsfox Team. Each guide is available as a full HTML page with sidebar table of contents, or as a printable PDF.
Beginner's Guide
First steps, core concepts, and your first GSD project. Ideal for users new to workflow automation and AI-assisted project management.
Intermediate Guide
Multi-phase projects, skill authoring, agent teams, chipset configuration, and production workflow patterns for experienced GSD users.
Advanced Guide
Deep configuration, custom chipsets, multi-agent orchestration, convoy operations, and performance tuning. For users scaling up to production.
Command Reference
Complete reference for all GSD v1.35.0 commands with syntax, flags, and examples. Covers all 75 canonical commands across every category.
Going Further
Customizing GSD in three deep walkthroughs: building custom skills and slash commands, integrating MCP servers (Stitch, Postgres, GitHub), and running multi-repository workspaces across several projects at once.
GSD-2 Architecture Research
GSD-2 Architecture — Research Series
Next-generation workflow system. Iron rule: every task fits in one context window. Disk-driven state machine. Milestone-Slice-Task hierarchy. Pi SDK integration. 7 research modules covering architecture, context engineering, extensions, git strategy, and verification.
GSD-2 Mission Pack
Complete mission package for GSD-2: LaTeX PDF, research index, integration document. Ready for execution via skill-creator.
Connection to Research
The GSD architecture is the tooling that builds this research site. Every page you read was planned, executed, and verified through GSD workflows. The chipset patterns map directly to distributed computing patterns studied in our WLCG research and formalized in the file-native agent architecture problem. The sweep cadence, mission packs, and convoy model are all GSD primitives in action.