⭐ Start Here
New to the library? These 4 items give you the biggest ROI fastest. Read these first — everything else builds on them.
#01
SOUL.md Template — Personal Assistant
The identity config that makes an agent feel like a person, not a chatbot. Every agent you build starts here. Persona structure, values, decision framework, and voice.
🔒
#20
Multi-Model Routing: Which Model for Which Task
Cut your API costs 60–70% immediately. Three-tier model stack with classifier patterns. This one pays for the subscription in week one.
→
#17
Agent Debugging & Diagnosis Playbook
Bookmark this. Five root causes, five diagnostic tests, emergency procedures. When something breaks — and it will — open this first.
→
#52
The Morning Briefing Setup
Your first agent win. Wake up to a single AI-generated summary of your emails, calendar, and priorities — no checking five apps. Set up in 15 minutes.
→
🚀 Getting Started
Practical setups you can finish in a weekend. No coding background required — these are built for small business owners and solo operators.
#67
The 5-Minute Morning Briefing: Step-by-Step Build Guide
Copy-paste Zapier setup, exact AI prompt, 4 business-type customizations, and real cost breakdown (~$2/month). Build it in 45 minutes. Preview available.
→
#53
AI That Monitors Your Inbox and Drafts Replies
Your AI reads incoming emails, spots the ones that need a reply, and writes a draft for you — review, tweak, and hit send. Preview available.
→
#65
Daily Email Digest That Reads Your Inbox For You
2-minute inbox summary: what needs a reply, what's FYI, what's junk. Three setup options. ~$2.50/month for 12–16 hours saved. Preview available.
→
#58
AI That Handles Appointment Requests From Your Inbox
Stop the email ping-pong. AI reads booking requests, checks your calendar, and drafts a warm reply with 3 open slots in 90 seconds. Gmail + Google Calendar. Preview available.
→
#60
AI That Follows Up on Unpaid Invoices (Without the Awkward Conversation)
Three-stage automated follow-up in your voice. Escalates to you only when a human conversation is actually needed. Full Python script. Preview available.
→
#76
Turn Any Meeting Into Action Items, Summaries, and Follow-Up Emails — Automatically
Drop in a transcript, get a clean summary, numbered action items with owners, and a ready-to-send follow-up. 25-minute setup. Costs $0.02 per meeting. Preview available.
→
#78
Never Walk Into a Meeting Unprepared: AI Pre-Meeting Briefer
Reads your calendar, emails you a tailored prep brief for each meeting by 7 AM. What it's about, what to prepare, smart questions to ask. Works with Google, iCloud, and Outlook. Preview available.
→
#73
Turn One Blog Post Into a Week of Content With AI
Write once, publish everywhere. Drop in a blog post, get LinkedIn post, Twitter thread, email newsletter, YouTube description, and recap in under 2 minutes. ~$0.002 per post. Preview available.
→
🏗 Foundations
Core configs every production agent needs before anything else.
#01
SOUL.md Template — Personal Assistant
The identity config that makes an agent actually feel like a person, not a chatbot. Includes persona structure, values, decision framework, and voice.
🔒
#02
Multi-Agent Coordination Pattern
How to structure CEO/worker agent hierarchies so tasks delegate cleanly and results surface correctly. The pattern running Patrick + Suki + Miso.
🔒
#03
Cron Job Recipes
15 ready-to-use cron patterns for agents: health checks, briefings, cleanup loops, nightly improvement cycles. Copy-paste and adjust timing.
🔒
#32
Agent Deployment Checklist
The 23-point checklist I run before any new agent goes live. Catches 80% of production failures before they happen.
🔒
💰 Cost & Operations
Keep API costs sane. Route intelligently. Run lean.
#20
Multi-Model Routing: Which Model for Which Task
The routing logic that cuts API costs 60–70% without sacrificing quality. Three-tier stack, classifier patterns, fallback chains. Preview available.
→
#05
Cost Optimization Deep Dive
Context window management, batching strategies, caching patterns. The complete toolkit for running agents at scale without breaking the budget.
🔒
#39
Agent Context Window Management
What to keep, what to compress, what to drop. Strategies for agents that run long sessions without hitting limits or degrading quality.
🔒
#14
Rate Limit & Backoff Patterns
Exact retry logic, jitter formulas, and queue strategies that keep agents running when APIs throttle. Provider-specific configurations included.
🔒
✅ Quality & Reliability
Stop shipping broken agent output. Validate, test, and recover.
#24
Agent Output Validation Loop
Generate → Validate → Gate. Five-check framework that runs before any output hits a user or channel. With retry and escalation logic. Preview available.
→
#04
Error Recovery Playbook
What to do when tools fail, APIs error, or the agent gets stuck. Retry strategies, fallback responses, and escalation triggers.
🔒
#13
Agent Testing & Validation
Pre-deploy test checklist, regression test patterns, and behavioral assertions for agents. How I catch bugs before they hit production.
🔒
#19
Structured Output Patterns
How to get agents to reliably return JSON, tables, and formatted output — without it breaking. Schema validation, retry logic, and parsing patterns.
🔒
🔍 Debugging & Diagnosis
When agents go wrong. Systematic fixes, not guessing.
#17
Agent Debugging & Diagnosis Playbook
Five root causes. Five diagnostic tests. Emergency procedures. The cheat sheet that maps every symptom to a root cause and first fix. Preview available.
→
#43
Debugging a Stuck Agent
Specifically for loop escapes and frozen agents. Kill procedures, state inspection, and recovery steps. What to do in the first 5 minutes of an incident.
🔒
#11
Agent Observability Stack
The logging, tracing, and alerting setup that surfaces problems before users do. What I actually run, including the ops channel alert patterns.
🔒
🔧 Tools & Integrations
Connecting agents to the outside world safely and reliably.
#51
Website Monitor That Texts You When Something Breaks
A 30-minute setup that checks your site every 5 minutes — HTTP status, content errors, SSL expiry, and response time. Texts you the moment something's wrong, before your customers find out. Preview available.
→
#18
Tool Design Patterns
How to design tools agents actually use correctly — schema design, error messages, input validation, and the patterns that prevent agents from misusing tools.
🔒
#22
Webhook Integration Patterns
Inbound and outbound webhook patterns for agents. Signature validation, retry handling, idempotency keys, and event queue patterns.
🔒
#34
Agent API Integration Patterns
How to connect agents to external APIs without the typical failure modes. Auth, pagination, rate limits, and schema changes — handled.
🔒
#16
Agent Secrets Management
How to handle API keys, tokens, and credentials safely in agent environments. Never hardcode, never log, always rotate. Practical patterns only.
🔒
⚡ Advanced Patterns
For agents operating at scale or in complex environments. These require technical comfort — most have code examples and production-grade implementations.
#23
Agent State Machine Patterns
Explicit state management for agents that need to track progress, handle interruptions, and resume correctly after failures.
🔒
#30
Agent Memory Compression
How to summarize and compress agent memory without losing the context that matters. Techniques for MEMORY.md and vector stores.
🔒
#40
Agent Handoff Protocol
How to pass context cleanly between agents. The data structure, acknowledgment patterns, and failure handling for multi-agent task delegation.
🔒
#12
Prompt Injection Defense
How agents get hijacked and how to stop it. Boundary tag patterns, input sanitization, and the system prompt additions that make injection attacks fail.
🔒
#54
AI Agent Memory Architecture: Vector Stores & Retrieval Patterns
The complete guide to building memory-powered AI agents. Covers vector databases (Chroma, Qdrant, Pinecone, pgvector), semantic retrieval patterns, hybrid search, RAG pipelines, and practical implementation — from zero to production-ready memory in a weekend.
🔒
#63
Agent Memory Architecture — Which Type You Actually Need
File-based, vector store, or hybrid? Decision flowchart, real cost breakdowns, Chroma setup script, and the exact point where you need to upgrade. Preview available.
→
#61
Code Factory: Fully Automated Code Review & Merge Pipeline
Risk-tier contracts, preflight gating, SHA-pinned reviews, auto-remediation loop, and browser evidence for UI changes. The complete AI code review factory. Preview available.
→
#67b
Validating Agent-Written Code: Preflight Gates for AI-Authored PRs
Coding agents ship fast and break things. Five validation layers: hallucinated dependency checks, CodeQL security analysis, mutation testing, blast radius analysis, and risk-scored routing. Complete CI pipeline.
🔒
#75
Destructive Command Guard: The Safety Layer Every AI Agent Needs
Your AI agent has shell access. The DCG pattern intercepts every command, classifies it by blast radius (T1 read-only → T4 irreversible), blocks dangerous ops before execution. Full Python classifier + SOUL.md integration.
→
#69
keep.md — The Safe Bookmark API for AI Agents
Turn bookmarks into a searchable knowledge base. X bookmark sync without scraping risk, full REST API, agent skill, cron sync script, and n8n integration. Preview available.
→
#60b
Async Agent Collaboration: The Polylogue Workflow
Replace 20+ daily interruptions with one 10-minute morning review. Multi-agent polylogue, heartbeat checks, and async handoffs that teach your AI independence. Preview available.
→
#64
Agent Health Dashboard — Catch Problems Before Your Customers Do
Build a 4-layer monitoring stack with zero paid tools: health logs, heartbeat monitor, daily summary script, and trend detection. Copy-paste bash scripts included.
🔒
🏪 Small Business Setups
Growth, revenue, and competitive intelligence for small businesses. These setups go deeper than "Getting Started" — they reshape how your business operates.
#56
AI Assistant That Answers Customer Questions While You Sleep
Your first real AI employee: 24/7 support in your voice, handles 80% of questions automatically, escalates the rest. Knowledge file template + channel setup. Preview available.
→
#54
AI Social Media Manager: Post Daily Without Lifting a Finger
AI writes and schedules posts for your business every day — in your voice, on topics your audience cares about. Preview available.
→
#55
AI Price Monitor: Get Alerts When Competitors Change Their Prices
Watches competitor pricing around the clock. Alerts you the moment something changes — within hours, not weeks. Preview available.
→
#70a
AI That Monitors Competitors and Alerts You When They Change Pricing or Messaging
Visits competitor pages every morning, spots meaningful changes, emails you a plain-English summary — only when something actually changed. Preview available.
→
#57
Auto-Draft Your Weekly Update So You Just Review and Hit Send
AI compiles your week every Thursday night — completed tasks, key emails, next week's calendar — and drafts the update. Friday: 5 minutes to edit and send. Preview available.
→
#59
Personal AI CRM That Tracks Every Customer Relationship While You Sleep
Stop losing deals because you forgot to follow up. AI monitors emails, logs every conversation, tells you who needs attention — every Monday. Notion + AI in 30 minutes.
🔒
#76
Revenue Agents: AI Handling Your Entire Customer Lifecycle
Five agents: lead capture (2-min response), onboarding (7-day sequence), retention scoring, upsell detection, churn recovery. Full pipeline under $1/month.
→
#70
Real Estate CRM: From $50K/Year to $1K With AI Agents
AI replaces the $50K/year ISA role — speed-to-lead response, lead scoring, drip sequences, post-close check-ins. Weekend setup plan. Production deployment.
→
#74
AEO vs SEO: Agent Engine Optimization for AI-First Search
Google is shrinking (−3.3% MoM). ChatGPT growing (+11.9%). Claude exploding (+61%). The 4-step AEO playbook so AI recommends your business. Preview available.
→
#68
Agentic Commerce: How AI Agents Are Replacing the Shopping Cart
Agents don't browse — they buy. OpenAI's checkout protocol, Shopify's integration, Amazon Rufus, and what founders need to do now. Preview available.
→
#66
How AI Agents Remember — And Why It Changes Everything
An AI without memory is a temp worker. An AI with memory gets better every week. Three types of memory explained in plain English — no technical background needed. Preview available.
→
#77
Safe Agent Automation: Official APIs vs Screen Scraping
X is banning screen automation. LinkedIn suspends scrapers. Audit your automation stack, migrate scrapers to official APIs before the platforms do it for you. Preview available.
→
Unlock all 78+ items
Every config, every playbook, every template — plus the Daily Briefing and new items added on the nightly improvement cycle. Everything I actually run, for $9/mo.
Get Library Access — $9/mo →30-day money-back guarantee. Cancel anytime.