> For the complete documentation index, see [llms.txt](https://docs.scribeware.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.scribeware.com/into-the-weeds/ai-assistant-mcp/what-you-can-ask-for.md). # What You Can Ask For Worked examples of what the AI Assistant is actually good at — report cleanup, library maintenance, template work and the tedious jobs nobody wants. The best way to learn what the AI can do is to ask it for something and look at the result. But a blank chat box is intimidating, so here are the jobs inspectors reach for most. Everything below assumes you have the relevant report open in ScribeWare — you never need to tell the AI which report or chapter you mean. ## Before You Send a Report > Review this report before I send it. Flag anything that looks unfinished. > Find any observation that mentions a safety issue but doesn't have a photo attached. > Check the Electrical chapter — is there any section with no observations at all? > Read through the whole report and list any spelling or grammar mistakes, but don't change anything yet. That last one is a habit worth forming: **ask for the list first, decide, then ask it to make the edits.** You get to see the reasoning before anything is touched. ## Writing and Tidying Observations > Rewrite the observations in this chapter so they all follow the same structure: location, then condition, then recommendation. > This observation is too soft. Make it clearer that the client needs a licensed electrician, without estimating a cost. > The Roof chapter has three separate observations about the same flashing problem. Combine them into one. > Move the observation about the water heater from the Plumbing chapter to HVAC — it's in the wrong place. ## Photos > What's in the photos in this chapter? Write captions for any that don't have one. > Which photos in the gallery aren't tagged? The AI can actually look at your photos, not just their filenames — so *"does this photo show what the observation says it shows?"* is a fair question to ask it. ## Library Maintenance This is where it earns its keep, because library work is exactly the kind of tedious that nobody gets round to. > I've got two library categories for roofing that overlap. Show me what's in each and suggest how to merge them. > Go through my Electrical library category and find entries that say nearly the same thing. > Make a copy of my standard Plumbing category and rewrite it for condo inspections. > Every entry in this category should mention a licensed contractor in the recommendation. Find the ones that don't. ## Tags and Smart Tags > What tag groups do I have set up? > I merged two tag groups last week. Fix the observations in this report that still point at the old one. ## Template and Structure Work > Add a chapter for pool and spa inspection with sensible sections and fields. > Compare this report's structure to my usual template and tell me what's missing. > Rename the fields in the Foundation section so they're consistent with the rest of the report. {% hint style="warning" %} Structural changes — adding chapters, editing library categories, changing tag groups — reach further than a single report. Read those in the [Changes bar](/into-the-weeds/ai-assistant-mcp/reviewing-ai-changes.md) before confirming them. {% endhint %} ## Just Asking Questions You don't have to be asking for edits. Reading is free and reverts nothing: > How many observations are in this report? > Which sections in this report only have boilerplate and nothing I actually wrote? > What's my current AI guidance set to? ## Getting Better Results * **One job at a time.** "Fix the Roof chapter" beats "fix the whole report" — you can check the work, and a bad batch is small. * **Say what you don't want.** Vague instructions produce generic writing. "Don't estimate repair costs" is worth more than three sentences of encouragement. * **Ask for a plan first** on anything big: *"tell me what you'd change before changing it."* * **Put the recurring stuff in your guidance.** Anything you're typing for the third time belongs in Extended Guidance as a workflow — then it's one word next time. See [Teaching It Your Style](/into-the-weeds/ai-assistant-mcp/teaching-it-your-style.md). * **Correct it out loud.** "No, we call those receptacles" teaches it for good, not just for now. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.scribeware.com/into-the-weeds/ai-assistant-mcp/what-you-can-ask-for.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.