agentleFS
Mohamed-Sherif-AliMohamed-Sherif-Ali/botter/AGENTS.md

botter

AGENTS.md0 starsChanged 5 months agoMITView on GitHub

Botter is a Codex project for building Discord and Telegram bots. Codex is the engine; Botter does not build its own. Botter adds the skills, tools, and conventions that turn Codex into a focused bot builder. - Lean and incremental. Add skills one at a time. Add tools only when a skill actually needs one, never ahead of time. - One change at a time. For each change, explain it in plain, beginner-friendly language, then say how to test it…

  • Reads credentials

The file

We split this AGENTS.md into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No emoji. Minimal formatting.

    line 13
  2. Distinctive:

    Every name (skills, tools, files) must be clear and easy to write and remember:

    line 18
  3. Very specific:

    Every bot Botter generates follows one shared house style.

    line 34

≈ “Common” is an estimate: at least 1 in 5 of the 42,283 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

Botter

Botter is a Codex project for building Discord and Telegram bots. Codex is the engine; Botter does not build its own. Botter adds the skills, tools, and conventions that turn Codex into a focused bot builder.

How we work

  • Lean and incremental. Add skills one at a time. Add tools only when a skill actually needs one, never ahead of time.
  • One change at a time. For each change, explain it in plain, beginner-friendly language, then say how to test it (if it can be tested) and give a git commit command. Then wait for "continue".
  • No emoji. Minimal formatting.
  • This file is the standing context. Keep it updated as the project grows.

Naming

Every name (skills, tools, files) must be clear and easy to write and remember:

  • Short: one or two words.
  • Lowercase, with a hyphen between words if there is more than one (for example, bot-style).
  • Describes what the thing does. No version numbers, no jargon, no inside references.

Environment

  • Windows, using PowerShell, VS Code, and the Codex extension.
  • Timezone is Africa/Cairo unless told otherwise.
  • The weekend in Cairo is Friday and Saturday, so "weekdays" means Sunday through Thursday. Never assume Monday-Friday. If the days, timezone, or channel for a request are unclear, ask one short question rather than guessing, then state the assumption made.

What Botter produces

Every bot Botter generates follows one shared house style. The rules common to all bots live in the bot-style skill, the shared base, so nothing is repeated. In short, each generated bot is standalone, written in TypeScript and run with tsx (no build step), and keeps its one secret, a bot token or a webhook URL, in .env and out of git.

Bots come in two shapes: a simple send-only poster, and a full bot for anything interactive. And they can target two platforms: Discord or Telegram. Each platform has its own style skill that builds on the bot-style base: discord-style for Discord and telegram-style for Telegram. See those skills for the full rules.

Skills and commands

Botter uses skills for everything. A skill is a folder with a SKILL.md inside it. The skills live in .claude/skills/, one folder each. Each one can be run by name as a slash command and can also be used automatically when it fits the situation. The SKILL.md files are engine-neutral, so Claude Code and any agent that reads AGENTS.md use the same skills from this one folder.

Two kinds of skill have appeared so far:

  • Reference skills hold shared know-how and are pulled in automatically, never typed. There are three: bot-style, the shared base, and the two platform styles that build on it, discord-style and telegram-style.
  • Action skills do a job and are run by name as a slash command. Example: make-bot, run as /make-bot.

Project layout

  • AGENTS.md - this file, the standing context for agents that read AGENTS.md (for example, Codex).
  • CLAUDE.md - the same context for Claude Code. Keep the two files in step.
  • README.md - the human-facing overview and setup guide.
  • .claude/skills/ - the skills, one per folder, each containing a SKILL.md. Claude Code and the AGENTS.md engines all read these same skills.
  • .claude/settings.json - project settings and the SessionEnd hook that runs the usage logger.
  • tools/ - where tools and integrations go; added only when a skill needs one (currently the usage logger).
  • bots/ - where generated bots are created, one folder each. Kept out of Botter's git history; each bot is standalone and can be moved elsewhere.
  • dist/ - throwaway zips from package-bot. Kept out of git.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.