Machine-Readable Documentation
AI agents and LLMs can access structured documentation for this project:
llms.txt— Curated index with descriptions and raw GitHub linksllms-full.txt— Complete documentation in a single file
These follow the llms.txt standard for AI-discoverable documentation. See the For AI Agents page for more.
Documentation Hub
Complete documentation for GSD Skill Creator, organized to mirror the llms.txt structure. Each entry links to both the site page and the raw GitHub source.
Getting Started
Quick Start Guide — Installation, prerequisites, first skill creation walkthrough (GitHub · Raw)
How It Works — End-to-end explanation of the observation-to-skill pipeline (GitHub · Raw)
Core Concepts — Skills, patterns, bounded learning, token budgets, and the coprocessor model (GitHub · Raw)
Configuration — All configuration options and environment variables (GitHub · Raw)
GSD Integration
GSD and Skill-Creator Overview — How the two systems complement each other (GitHub)
GSD Orchestrator — Natural language routing to GSD commands via the 5-stage classification pipeline (GitHub · Raw)
Skills & Learning
Skill Format — Markdown skill file structure, frontmatter fields, activation triggers (GitHub · Raw)
Pattern Discovery — How the system detects repeating tool sequences and user corrections (GitHub · Raw)
Bounded Learning — Safety guardrails: refinement limits, cooldowns, user confirmation (GitHub · Raw)
Token Budget — 6-stage skill loading pipeline and context window management (GitHub · Raw)
Agents
Skills vs Agents vs Teams — How skills compose into specialized agents from co-activation patterns (GitHub · Raw)
Agent Teams — Multi-agent orchestration and team coordination patterns (GitHub · Raw)
Reference
CLI Reference — All CLI commands, flags, and usage examples (GitHub · Raw)
API Reference — Programmatic API for embedding skill-creator in other tools (GitHub · Raw)
Extensions — Available extension points and third-party plugins (GitHub · Raw)
File Structure — Project directory layout and file purposes (GitHub · Raw)
Architecture
Architecture Overview — High-level system architecture (GitHub · Raw)
Layer Architecture — The layered design from storage to orchestration (GitHub · Raw)
Data Flows — How data moves through the system (GitHub · Raw)
Tutorials
Creating Your First Skill — Step-by-step guide to skill creation (GitHub · Raw)
Creating Agent Teams — Setting up multi-agent teams (GitHub · Raw)
Calibrating Thresholds — Tuning activation thresholds and confidence levels (GitHub · Raw)
CI/CD Integration — Adding skill-creator to CI/CD pipelines (GitHub · Raw)
Detecting Conflicts — Handling skill conflicts and resolution strategies (GitHub · Raw)
Troubleshooting
Troubleshooting Guide — Common issues, error messages, and fixes (GitHub · Raw)
Development Setup — Contributing, local development setup, running tests (GitHub · Raw)
Requirements — System requirements and compatibility matrix (GitHub · Raw)
Source Code Browser
The codebase is documented across 50 module pages in the Source Code Browser. Each module page covers a specific src/ directory with type definitions, key functions, and architectural notes extracted from the real TypeScript source.
Quick Links
- Quick Start Guide — Get up and running in 5 minutes
- Release Chronicle — 33 versions of development history
- About the Project — Philosophy, bounded learning, and project overview
- GitHub Repository — Source code, issues, and contributions
This documentation hub mirrors the structure of llms.txt and llms-full.txt, which provide machine-readable versions of this documentation for AI agents and LLMs.