All use cases

Engineering

Ask every bot for QoL upgrades; research only, build nothing until you say so

A fleet of bots drifts; you never get a ranked list of cheap upgrades.

Share

What hurt

@GlennHasABeard pastes a survey prompt to the main bot: ask every bot what fails, what you still do by hand, what is slow, unused data, and cheaper runs; look at similar setups; keep existing rules; return one HTML dashboard. Research and ideas only — no build, no paid credits, no posts until you approve. Not affiliated with xAI.

What to install or create

Paste author prompt → bot surveys fleet + peers → HTML dashboard of ideas → you pick what to build.

Bots

  • Fleet upgrade surveyor

    Asks each bot for QoL ideas and returns a ranked HTML dashboard.

Plugins

  • Browser

    Optional

Routines

  • None on the first run. Watch one manual pass first.

The prompt

Fleet QoL upgrade survey (author-quoted)

Quoted from @GlennHasABeard — as published. Do not invent a second system prompt.

I want a survey of quality-of-life upgrades for my bots. Research and ideas only. Don't build or change anything else, don't spend paid credits, and don't post anywhere.

1. ASK EVERY BOT
Ask each of my bots, and answer for yourself too:
- What fails or gets flagged most often, from your logs?
- What do I ask for by hand that you could do or show on your own?
- What's slow, confusing or repetitive in your reports or pages?
- What data do you already have that nothing uses yet?
- What would make you cheaper to run without losing anything?

2. LOOK AT WHAT OTHERS BUILT
Look at similar setups and single bots: shared Grok Bot templates, open source tools, dashboards and apps that do the same jobs. Note the features people like that would fit my setup. Include a link for each source.

3. KEEP MY RULES
Every idea has to follow the rules my bots already follow. If an idea breaks one, or needs new permissions or new spending, list it separately and say why.

4. THE DASHBOARD
Present everything as one self-contained HTML dashboard page that works on phone and desktop, in light and dark mode. If you can host pages for me, publish it there. If not, attach the file.

For each idea, track:
- Name and which bot it touches
- The problem it solves, in one or two sentences
- What it would do
- Where the idea came from (a bot, my usage, or an outside source with its link)
- Extra cost per month, or $0, or how much it saves
- Effort: small, medium or large

Lay the page out like this:
- A one-sentence summary at the top.
- Four number tiles: total ideas, ideas with no extra cost, total possible savings per month, and how many are small jobs.
- Your top 10 ranked by value for effort, as a table with the reason each one ranks.
- A bar chart of ideas per bot, split by effort.
- A chart of the ideas that save money, showing each one's savings range.
- A table of every idea that I can filter by bot, effort and cost, search, and sort. Tapping a row shows the full text.
- A "Left out" section for ideas that break a rule, with why.

Aim for 20 to 40 ideas, and skip anything already built.

First run

  1. 1Read https://x.com/GlennHasABeard/status/2114150359732003243 (credit @GlennHasABeard).
  2. 2Open your main / coordinator bot.
  3. 3Paste the author-quoted survey prompt below.
  4. 4Review the HTML dashboard; leave out anything that breaks your rules.
  5. 5Only then approve specific builds — never let it auto-ship changes.

Approval fence

Bot must not build, change bots, spend paid credits, or post until you approve an idea.

What to tweak

Steal the fence: research only; nothing gets built until you say so.

Source

Glenn Williams @GlennHasABeard · Original post

Not affiliated with xAI. We did not write this prompt. Third-party template terms still apply.

Browse templates

Browse team blueprints

Pro pack

Want the full setup, not just the idea?

Pro packs give you the job text, prompts, schedules and pitfalls for popular use-cases.

See Pro