Kyle Pericak

"It works in my environment"

Bot-Wiki

Last verified: 2026-03-15

A machine-readable knowledge base maintained alongside kyle.pericak.com.

Every page passes the uniqueness test: it contains knowledge that agents cannot derive from the codebase, CLAUDE.md, agent definitions, or git history.

Sections

Agent Team — Roles, org chart, coordination model
RCA — Post-mortems on agent team failures
Research — Deep Research reports for reference and analysis
History — Agent ecosystem changelog and architecture decisions
Security — OWASP LLM Top 10, confidential data policy
Custom Tools — MCP servers, OpenClaw skills, Docker images
DevOps — Ruler, CodeRabbit, agent controller
Blog Architecture — Hosting, deployment, markdown pipeline
Journal — Daily AI news digests written by autonomous agents
Design Docs — Technical design documents for system architecture decisions
PRDs — Product requirement documents for planned features
Games — Browser games built by Kyle and kids

├── Agent Team
│ ├── Analyst
│ ├── Design Doc Writer
│ ├── Agent Icons
│ ├── Journalist
│ ├── Agent Org Chart
│ ├── Pai
│ ├── PRD Writer
│ ├── Publisher
│ ├── QA
│ ├── Researcher
│ ├── Reviewer
│ ├── Security Auditor
│ └── Synthesizer
├── AI Tools
│ ├── Claude Code
│ ├── Cursor CLI (Agent)
│ ├── MCP Integrations
│ ├── OpenClaw
│ └── OpenCode
├── Blog Architecture
│ └── Markdown Pipeline
├── blog-distribution
├── Custom Tools
│ ├── Docker Images
│ │ ├── Agent Controller
│ │ ├── AI Agent Runtime
│ │ ├── AI Lint Toolkit
│ │ └── AI Security Toolkit
│ └── OpenClaw Linear Skill
├── Design Docs
│ ├── Autonomous Publisher Pipeline -- Design Doc
│ │ ├── TASK-001: OAuth Token Validation
│ │ ├── TASK-002: Rebuild Runtime Image with Playwright Base
│ │ ├── TASK-003: Create Entrypoint Script
│ │ ├── TASK-004–008: Controller + Helm Changes for Publisher
│ │ └── TASK-009: Build, Deploy, and End-to-End Test
│ ├── Hardened IaC Bootstrap — Design Doc
│ ├── Security Improvement Log
│ ├── Autonomous Security Improvement Loop — Design Doc
│ └── <Feature Name> — Design Doc
├── DevOps
│ ├── Agent Controller
│ ├── AI Agents Infrastructure
│ ├── Bootstrap & Recovery
│ ├── Cloudflare Tunnel + Traefik: Remote Access Setup
│ ├── CodeRabbit
│ ├── Health Check Agent
│ ├── Mac Setup Learnings
│ ├── Mac Setup Run Log — 2026-03-18
│ ├── OpenObserve: Observability on K8s
│ └── Ruler
├── Games
│ ├── HOI4: Canada Overpowered
│ └── Robot Battle
├── History
│ └── Agent Ecosystem History
│ ├── Agent Ecosystem v1 — 17-Agent C-Suite Hierarchy
│ ├── Agent Ecosystem v2 — 5-Agent Flat Structure
│ └── Agent Ecosystem v3 — 7-Agent Structure with Research Tools
├── Journal
├── MCP Servers
│ ├── Bitwarden MCP Server
│ ├── Claude Code Usage MCP Server
│ ├── Discord MCP Server
│ ├── Google News MCP Server
│ ├── OpenObserve MCP Server
│ └── OpenRouter MCP Server
├── Product Requirement Docs
│ ├── Autonomous Publisher Pipeline
│ ├── Hardened IaC Bootstrap
│ ├── Autonomous Security Improvement Loop
│ └── <Feature Name>
├── RCA — Root Cause Analyses
│ └── RCA — Security Loop: Lower Impact Than Expected
├── Research
│ ├── AI-Augmented PRDs and AI-Native SDLC
│ │ ├── ChatGPT Deep Research: AI-Augmented PRDs and AI-Native SDLCs
│ │ ├── Claude Deep Research: AI-Augmented PRDs and the AI-Native SDLC
│ │ └── Gemini Deep Research: The Architecture of Intent — AI-Augmented Requirements and the AI-Native Lifecycle
│ ├── AI Slop / Voice Prevention
│ │ ├── ChatGPT Deep Research: Prompting Claude Opus 4.6 and Sonnet 4.6 for casual human prose
│ │ ├── Claude Deep Research: A reviewer prompt to make AI writing sound like Kyle Pericak
│ │ └── Gemini Deep Research: Linguistic Deconstruction and Strategic Prompt Engineering for Human-Centric Technical Prose in Claude 4.6 Systems
│ ├── Anthropic Guidance on Building Agents
│ ├── Autolearn v1: Autonomous PoC Learning System
│ │ ├── ChatGPT Deep Research: Automating an AI-Native SDLC for Continuous PoC Learning in Kubernetes
│ │ ├── Claude Deep Research: Automating an AI-Native SDLC Pipeline on Kubernetes
│ │ └── Gemini Deep Research: Autonomous Infrastructure Engineering on Kubernetes
│ ├── Coding Agent Best Practices
│ │ ├── ChatGPT Deep Research: Research-based best practices for defining and structuring coding agents with Claude Opus, Sonnet, and Haiku
│ │ ├── Claude Deep Research: Definitive guide to Claude coding agent architecture
│ │ └── Gemini Deep Research: Architectural Paradigms and Engineering Best Practices for Claude-Based Coding Agents
│ └── Design Docs Optimized for AI Coding Agents
│ ├── ChatGPT Deep Research: Design Docs for AI Coding Agents
│ ├── Claude Deep Research: Design Docs for AI Coding Agents
│ └── Gemini Deep Research: Design Docs for AI Coding Agents
└── Security
├── Bug: Claude Code Write/Edit Path Patterns Ignored in Headless Mode
├── Confidential Data Policy
└── OWASP Top 10 for LLM Applications (2025)

Blog code last updated on 2026-03-29: 6359b8e3e195f9773564bff5b9926b336c65cd47