AI Fluency Hub
Infographics, templates, and prompt packs – published weekly.
From Build Queues to Bet Registers: Replacing Your Roadmap with a Hypothesis Portfolio
A roadmap is a build queue. A bet register is a portfolio of hypotheses. Learn how to replace fixed delivery plans with structured experiments that uncover what users actually need.
How to Build Android Apps with Google AI Studio
Google AI Studio now lets you build Android app prototypes and features using natural language. Here is how product managers can leverage it to accelerate discovery, validate ideas, and ship faster.
Webwright: A Terminal-Native Web Agent Framework That Scores 60.1% on Odysseys
Microsoft's Webwright is a terminal-native framework that turns any web browser into an AI agent. It scores 60.1% on Odysseys, nearly doubling base GPT-5.4's 33.5%. Here is how it works and how to use it.
How the Hermes Agent Streamlines Daily PM Workflows
The Hermes Agent automates repetitive PM tasks like stand-ups, status updates, and backlog grooming. Here's how to integrate it into your daily routine.
How to Pick the Right LLM for Every Task (and Save Money)
Using the most expensive model for every query is like using a sledgehammer for a thumbtack. Learn to match models to tasks and slash your AI costs without sacrificing quality.
Three No-Code Automations You Can Build with Claude in an Hour
Stop waiting for engineering resources. Learn to build simple, powerful automations like Slack alerts and email digests yourself using Claude Code, no experience required.
From Vague Idea to Working Code: The 5-Part Prompt Structure
Getting unexpected results from Claude? The issue is rarely the model, but the prompt. This five-part structure provides the clarity Claude needs to build what you actually envision.
Beyond Blind Trust: A Pre-Flight Checklist for Claude Code
Stop blindly trusting Claude Code. Senior engineers use a pre-flight checklist to verify context, constrain scope, and set review checkpoints. This ensures AI accelerates, not derails, your work.
How AI Agents Can Supercharge Product Management
Stop drowning in busywork. A personal AI agent can automate competitive analysis, user story generation, and stakeholder comms, freeing you to focus on high-impact strategy.
How to Structure Your CLAUDE.md for a Product Team
Most CLAUDE.md files are one-paragraph boilerplates that Claude ignores within the first message. This resource gives you the five-section structure that makes Claude a functional PM teammate: product identity, user definitions, current sprint goal, settled decisions, and working principles. Includes a ready-to-copy template with annotations.
How to Describe What You Want So Claude Builds the Right Thing
The gap between what you described and what Claude built is almost always a description problem. This resource gives non-technical builders a five-part prompt structure - context, goal, constraints, format, and examples - that closes that gap. Walk through each element with before-and-after examples so your next Claude session produces what you actually meant.
How Builders Use Sub-Agents for Parallel Research Tasks
Most agentic setups break down when multiple agents work on the same problem at once. This resource shows how to structure parallel sub-agent tasks so each agent has a clear scope, a defined output format, and a handoff the next agent can consume cleanly. Includes the coordination pattern that keeps outputs coherent without constant manual review.
The Pre-Prompt Checklist for PMs
Most PMs jump into a Claude session without loading the context that shapes the output. This checklist covers five things to encode before every session: product identity, user definitions, sprint goal, settled decisions, and working principles. Load these once and the first response is already calibrated to your situation.
How to Write a PRD with Claude Code Without Losing Your Voice
Most PRDs written with AI sound like they were written by AI. The gap is context — Claude does not know your product's language, your team's mental models, or how you frame decisions for stakeholders. This resource shows you how to encode that context in your CLAUDE.md so every PRD Claude drafts sounds like you wrote it.
From User Interview to Opportunity Statement in One Session
The gap between a user interview and a sprint-ready opportunity statement is where most discovery work gets lost. This resource gives you the exact prompt sequence to run in a single Claude session — from raw transcript to structured opportunity statement — including how to handle contradictory signals and what the output should look like before sprint planning.
Competitive Intelligence Automation for PMs
Most PMs spend hours each week manually scanning Reddit, competitor changelogs, and social feeds — only to produce a brief that is already stale. This resource gives you the exact Claude setup to automate your weekly competitive scan: source list, loop structure, and output format that makes the brief actually useful to your team.
Running a Discovery Sprint with Claude Code
Discovery sprints generate dozens of browser tabs and documents that nobody reads. This resource shows PMs how to run an entire sprint — interview synthesis, pattern clustering, and opportunity statements — in a single Claude Code session. You leave with structured output that feeds directly into sprint planning.
The 7 Claude Code Commands Every PM Should Know
You do not need to memorize every Claude Code command. You need to know seven. Here is what each one does and when to use it as a PM.