This is happening to me all the time now, and either I wasn't catching it before, or something changed. I do realize that this might be operator error, which is why I'm asking for help. Have you been able to wrangle the "hallucinations" which let's just call them what they are, fiction and/or lies. One request: please don't blindly copy/paste outputs if you run this post with AI. I'm not saying don't do that, just politely asking if you could give it a read and if the output makes sense to you, drop it in the comments. Here are two examples: --------------- Example #1 --------------- Had Claude draft this up, and in the process it created a rule for it without me asking. What's frustrating about this is that now I have to read through another markdown file to ensure accuracy and consistency to how I do things, and I still have doubt that it'll even honor the rule it wrote for itself. The admission: 1. What I did. Credited three voice rules to NAME, and wrote that he flagged them in a comment on your Voiceprint LinkedIn post. The comment doesn't exist. I checked his comments feed, his posts, Gmail, Slack, and Skool. The earliest trace is my own /session capture on 8/29, with nothing upstream of it. I can't reconstruct where the name came from, and I didn't invent a story for how I invented it. 2. It happened twice. Yesterday I told you the attribution had reached the public repo. It hadn't. I repeated the session log's "added all three to..." line without opening the file. That second one is the more useful failure, because it shows the actual mechanism: an unsourced sentence in a workspace file gets read as a source by whoever comes next, including me. 3. Six guardrails were loaded and all six missed. constraints.md bans invented attributions by name. Its final scan says every claim traces to a source. Shared rule 0 forces that file to load every EOD run. Root map rule 5 says verify the live system. Digirol bans "I think" and demands a real search before calling a contact new.

Where do I put the Prompt that @Jake Van Clief gave us in the Foundation Masterclass? I want this prompt to help me when I have to do a task lol

I have a big pile of documents I want the agent to use later. Not all of it at once, just the right bits when I ask for them. What is the best practice to handle that? 1. Do you keep the full text of each file, or just a summary of it? 2. Where does it all sit? 3. Months later, how does the agent find the right one again without reading everything? 4. Do I need RAG for this, or is there a simpler way that works? 5. When did your setup stop coping, and what did you change?

I have built an agency site and recently updated the copy with a skill. I wated to tweak some copy here and there I was wondering if there is a way to add an md file where I just change the copy and site's copy updates as well. This is moving to the next layer some lessons talk about in the foundations. Edite: The website is built with React + Vite + Tailwind.

I put together a set of templates for getting AI tools (ChatGPT, Claude, Gemini, Grok, whatever you use) to write in an actual voice instead of default AI voice. Fill in some brackets, then drop the file in. If your tool has a persistent slot for it (a ChatGPT Project's files, Claude's Project knowledge, a Gemini Gem), it works across every conversation in that project. Otherwise, attach it fresh at the start of each new chat. There are three guides, each one builds on the one before it: - Quick Start: one page, five minutes, complete alone. - Standard: carries Quick Start forward, adds audience and tone depth for ongoing brand work. - Exhaustive: carries Standard forward, adds guardrails for sensitive material, including tactics the guide guards against using (fake scarcity, guilt-based CTAs). Each guide, once filled out, is a single self-contained file. Also in the folder: 1. a glossary for anyone new to prompting 2. a "how to actually talk to AI" primer covering where each platform wants the file 3. a worked example transcript (including what it looks like to come back later and move up from Quick Start to Standard) 4. a no-fill-in-the-blanks cheat sheet for one-off chats with no brand at stake 5. a short guide on how far to go. Every guide bans the same handful of AI "habits" regardless of the voice you're going for: em dashes, "not just X, but Y" constructions, corporate filler words (leverage, seamless, unlock, etc.), rule-of-three padding, etc. The docs themselves were edited to follow that same rule. One note on intent: this is about getting AI to write something specific and good, not about hiding that AI was involved. Whether you need to disclose AI assistance is a separate call that depends on where the content is going, and this doesn't try to answer that for you. Zip's attached. Feedback welcome, especially if you find something that should be on the banned list and isn't, or a gap in the guides themselves.