Claude Code Mentoring
Your team is using 10% of what's available. I'll show them the other 90%.
Most teams are using A.I. tools the way they used Google in 2004
They type a question, get a bunch of code, hope it works, and move on. That’s the simplest possible interaction with the most powerful general-purpose tool released in a generation. Claude Code—Anthropic’s agentic coding and research environment—can write production code, execute multi-step research, manage files, build pipelines, and orchestrate complex workflows across your entire stack. But most teams treat it like a chatbot. The difference between a team that uses Claude Code at 10% and one that uses it at 90% isn’t raw intelligence—it’s technique.
Prompting patterns, tool composition, memory management, skill injection, and workflow design. These are learnable, teachable disciplines. I’ve spent every day for the last 9 months in Claude Code, building tools that make it more accessible to developers and non-tech creators as well.
What I do
This is hands-on pair programming, not a lecture series. Your team builds their own workflows during our sessions, on their own codebases, with their own data. I consult alongside them.
- 1-on-1 onboarding—A focused 1–2 hour session tailored to your role and industry. I’ve onboarded 20+ people this way—marketers, designers, developers, product managers—each walking away with Claude Code running on their own work.
- Prompting architecture—Move beyond “ask a question, get an answer.” Learn structured prompting patterns that produce reliable, reproducible output across sessions.
- Skill and tool composition—Build custom Claude Code skills, integrate hooks, and compose tools that extend Claude’s capabilities to match your specific workflows.
- Memory and context management—Design persistent memory systems (CLAUDE.md, project instructions, handoff protocols) so your team’s institutional knowledge survives across sessions and across engineers.
- Workflow orchestration—Chain multiple Claude Code sessions, spawn sub-agents for parallel work, and build review/iteration loops that ship production-quality code.
- Research and analysis—Use Claude Code as a genuine research partner—web scraping, data extraction, competitive analysis, literature review—not just a code generator.
Open-source foundations
I don’t just teach Claude Code—I build and open-source the frameworks that push it further. Claudicle is a 4-layer soul agent framework I built for persistent identity, memory, and cognitive architecture across Claude Code sessions. The Claude Code Minoan skill ecosystem extends Claude with 60+ specialized skills—from web scraping to academic research to image generation.
These are the same tools and patterns I teach in mentoring sessions—battle-tested in production, not theoretical. Even smaller tools like Dabarat (8 stars)—a zero-dependency Markdown previewer with annotations and bookmarks—were built entirely with Claude Code in a single session.
Who this is for
Engineering teams who want to ship faster without hiring. Product teams who want to prototype in hours instead of sprints. Marketing and ops teams who want to automate research, reporting, and content workflows. Founders who are building solo and need to multiply their output. If your team touches a keyboard, Claude Code can make them faster.
What you get
- Custom training sessions (2–4 hours) tailored to your team’s stack and workflows
- Written playbook of prompting patterns, skills, and configurations specific to your use case
- CLAUDE.md and project instruction templates for your codebase
- Follow-up session (included) to review what your team has built and troubleshoot
- Ongoing Slack or email support for 30 days post-training
The teams that learn to work with A.I. tools at depth—not just at the surface—will outpace their competitors by a margin that compounds every quarter.
Claudicle 12 sub-daimones for specialized reasoning—research, reflection, design, verification, memory retrieval. Each runs as a focused subagent with constrained tools.
claude-code-minoan 0Three-tier persistent memory across terminal, Slack, and SMS channels. Working memory, soul memory, and user models survive across sessions and machines.
Terminal, Slack Socket Mode, SMS via Telnyx/Twilio, REST API. Each channel adapter maps to the same soul engine with channel-specific formatting.
dabarat 0See it in action
A real Claude Code workflow—from repo audit to deployed skill in three commands:
Ready to start? Get in touch today.
The first call is free. 30 minutes, no pitch—just your stack, your problems, and what I'd do about them. Book a call.