agentleFS
Mohamed-Sherif-AliMohamed-Sherif-Ali/botter/.claude/skills/bot-style/SKILL.md

bot-style

Skill0 starsChanged 5 months agoMITView on GitHub

The shared base style every Botter bot follows, on any platform. The platform style skills (discord-style and telegram-style) build on this; read it first for the rules common to all bots. Covers the standalone layout, TypeScript with tsx, the secret-only .env, Cairo scheduling, saving data to be host-ready, and the two shapes (a send-only poster and a full bot) with when to use each.

  • Reads credentials

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    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:

    Keep it minimal and beginner-friendly.

    line 24
  2. Distinctive:

    Every bot takes one of two shapes.

    line 28
  3. Very specific:

    Keep the data in a simple JSON file inside the bot's folder (for example, data.json).

    line 56
  4. Distinctive:

    Write saves safely: write to a temporary file, then replace the real one, so a crash mid-save cannot corrupt the data.

    line 59

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill 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
SKILL.md

bot-style

The shared base style for every bot Botter generates, whatever the platform. These are the rules common to all bots. Each platform then has its own style skill that builds on this one: discord-style for Discord and telegram-style for Telegram. Read this skill for the shared rules, then read the matching platform skill for the specifics.

Shared rules (every bot)

  • Standalone and runnable on its own: its own folder, its own dependencies, its own .env. Someone should be able to copy the folder, add one secret, and run it.
  • Node.js with TypeScript, run with tsx. No build step.
  • dotenv loads the one secret from .env. croner handles anything time-based, always with an explicit timezone (Africa/Cairo by default) so schedules never depend on the machine's clock.
  • The one secret lives only in .env. Never hardcode it. .env is git-ignored, and a .env.example with no real value shows what to fill in. .gitignore also lists node_modules.
  • Cairo locale: the weekend is Friday and Saturday, so "weekdays" means Sunday through Thursday. Never assume Monday-Friday.
  • Keep it minimal and beginner-friendly. Don't add features nobody asked for.

The two shapes, and when to use each

Every bot takes one of two shapes. Always choose the simplest that does the job. The platform style skill says how each shape is actually built; this is just the choice between them.

  • Poster - a send-only job: it only posts messages, such as reminders, announcements, scheduled posts, or status updates. Simpler, and all most bots need. It cannot read, respond, or react.
  • Full bot - use when the job needs more than posting: responding to people, reading messages, reacting, handling commands, or working across more than one place at runtime.

Rule of thumb: if the job is only sending messages, use the poster; if it must listen, use the full bot.

Suggested project layout

my-bot/
  src/index.ts       the entry point
  package.json       dependencies and a "start" script that runs src/index.ts with tsx; private
  tsconfig.json      modern, sensible settings
  .env               the one secret (never committed)
  .env.example       shows the variable needed, with no real value
  .gitignore         lists .env and node_modules
  README.md          plain run steps

Saving data, and being host-ready

Some bots accumulate data while they run: a tracker counting time, a counter, a leaderboard. That data lives only in memory, so a crash, a reboot, or a move to a host would wipe it. If a bot remembers anything across time, it must save that data to a file and load it again on startup.

  • Keep the data in a simple JSON file inside the bot's folder (for example, data.json). Load it when the bot starts, and save it whenever it meaningfully changes and right after producing any scheduled report.
  • Write saves safely: write to a temporary file, then replace the real one, so a crash mid-save cannot corrupt the data.
  • This file is the bot's own data, not a secret, but it should not be committed or shared. List it in the bot's .gitignore, and leave it out when packaging the bot so the recipient starts fresh.
  • On startup, carry on cleanly from whatever state was loaded, and account for anything already in progress (for example, members already sitting in a voice channel when the bot starts).

This is also what makes a bot host-ready: together with the standalone layout, the secret-only .env, and an explicit timezone, a bot that saves and reloads its data can move to an always-on host and survive restarts there without losing anything.

Guiding the user through setup

Whatever the platform, do every step that can be done for the user, and hand-hold the few that cannot, in plain, non-technical language. Never ask for anything technical the user does not need: no IDs, no permission numbers, no hand-built URLs. Always, once the user pastes their secret, write it into .env for them, run npm install, and give the exact run command. The platform style skill covers the platform-specific setup steps.

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.