21 battle-tested patterns for building and operating AI coding agents. From parallel sessions to fleet coordination, from routing to memory — available as PDF or live through your coding agent via MCP.
Conduit turns your everyday browsing into structured, actionable data — for you, your spreadsheets, or your AI.
Conduit silently captures the data flowing through your browser and serves it to AI agents, scripts, and dashboards via MCP + REST.
The definitive Agent DevOps OS (ADO) for guaranteeing deterministic behavior from unpredictable AI coding agents.
A 100% Bash-native Git Submodule that acts as a secure "Bridging Adapter" between non-deterministic AI and your fragile repositories.
21 battle-tested patterns for building and operating AI coding agents — from workflow templates to cognitive architecture.
Everything you need to run AI coding agents in production — and build the cognitive systems behind them. 21 chapters in two parts: DevOps patterns (how to operate) + architecture deep-dives (how to build).
One MCP server — 21 chapters of agent architecture + DevOps patterns. 3 free chapters included.
One MCP server, 21 chapters — architecture deep-dives AND DevOps patterns
List all 21 Handbook chapters. Two parts: Agent DevOps Patterns + Architecture Deep Dive.
Read any chapter by number (0-20) or topic keyword. Rich markdown with code examples.
Semantic search across all 21 chapters. Returns the most relevant sections with scoring.
Look up a specific architectural pattern with implementation details and code.
21 chapters · 91 pages · 22K+ words · 4 MCP tools
From a system running 24/7 in production