← Back to catalog

Elie Steinbock launches Bot Directory for one-prompt Grok Bots

Public plays become one catalog. GitHub is the store. Tag a tweet and it can open a PR.

For

You want to see how other people run this, or send a play you already ran back in.

When you're done

You can filter the catalog by job. Your play goes in by PR, or by tagging their account.

How

  1. 01
    Open botdirectory.ai.

    Filter by job or tool. Pick one you will actually run this week. Stand the bot up from its job.

  2. 02
    Run it first. Then save.

    Do not treat the directory as a bookmark pile.

  3. 03
    To send your own play back, open a PR.

    The file has the name, the category, who you are, what you used, and how it works.

  4. 04
    Or tag them under a tweet that describes the setup.

    Their agent reads the post and opens a PR. Run it end to end yourself first. Spend and outbound send go in the bio.

Notes

What's good: you get plays other people already ran, not a pile of empty role names. The catch: the directory is a shelf. If you do not run it, pasting it does nothing.

Prompt

第一条消息:按目录挑一条并建 Bot(不是目录代理岗位)
I want to use Bot Directory as a catalog, not invent a workflow from scratch. Help me pick ONE listing on botdirectory.ai for a job I will actually run this week, then stand that listing up as a named Grok Bot. Walk me through the connectors it asks for — I will sign in on your computer, not in chat. Do a test run, then save it as a bot I can edit. If I later contribute my own play, we will either open a GitHub pull request (name, category, contributor, integrations, how it works) or I will reply under an X post that describes the setup and tag @BotDirectoryAI so their agent can file the PR. I will only submit something I have already run end to end. Do not invent a Bot Directory agent job; the directory is markdown files plus that intake loop.

改写自 https://runtimewire.com/article/elie-steinbock-launches-bot-directory-grok-bot-prompts

Related plays

Write-ups follow public posts and official docs. No likes, views, or dashboard numbers. Open the source to check the original. About