agentleFS
tjg37tjg37/ironclaw/.claude/skills/setup/SKILL.md

setup

Skill1 starsChanged 7 months agoMITView on GitHub

Interactive setup for IronClaw — checks prerequisites, starts services, configures auth and agent persona

  • Reads credentials

192 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use add-voice-transcription, add-discord and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • juancruzmunozalbeloSkill

    add-voice-transcription

    Add voice message transcription to NanoClaw using OpenAI's Whisper API. Automatically transcribes WhatsApp voice notes so the agent can read and respond to them.

    Also in 5 projects that use this

  • MicrosoftSkill

    add-discord

    Add Discord bot channel integration via Chat SDK.

    Also in 11 projects that use this

  • MicrosoftSkill

    add-slack

    Add Slack channel integration via Chat SDK.

    Also in 11 projects that use this

  • juancruzmunozalbeloSkill

    add-telegram-swarm

    Add Agent Swarm (Teams) support to Telegram. Each subagent gets its own bot identity in the group. Requires Telegram channel to be set up first (use /add-telegram). Triggers on "agent swarm", "agent teams telegram", "telegram swarm", "bot pool".

    Also in 7 projects that use this

  • +8 more from microsoft

The file

Show its rules and text

We split this skill into the rules it gives an agent

26 rules. 0 are common, 8 are distinctive, and 18 are very specific to this project.

  • 0

    Common

    Most of the 169 edited versions we read keep these.

  • 8

    Distinctive

    This author's own choices. Worth a look.

  • 18

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    IMPORTANT: Use the AskUserQuestion tool for ALL user choices.

    line 10
  2. Very specific:

    Check Node.js version: run node --version

    line 14
  3. Very specific:

    If Node.js is missing or below v22:

    line 15
  4. Very specific:

    If nvm is not available, tell the user: "Node.js 22+ is required.

    line 17
  5. Very specific:

    Stop setup if Node.js 22+ cannot be activated

    line 18
  6. Very specific:

    Verify pnpm is available (pnpm --version).

    line 19
  7. Very specific:

    Verify Docker is running (docker info)

    line 20
  8. Very specific:

    If Docker is not running, tell the user to start Docker Desktop and re-run /setup

    line 21
  9. Very specific:

    Run pnpm install from the project root

    line 25
  10. Distinctive:

    Verify it completes without errors

    line 26
  11. Very specific:

    Check if Docker Compose services are already running: docker compose -f docker-compose.dev.yml ps

    line 30
  12. Very specific:

    If not running, start them: docker compose -f docker-compose.dev.yml up -d

    line 31
  13. Distinctive:

    Wait for Postgres to be healthy

    line 32
  14. Distinctive:

    Use AskUserQuestion (single select) with:

    line 37
  15. Very specific:

    Run claude auth status to verify they're authenticated

    line 42
  16. Very specific:

    If not authenticated, tell them to run claude auth login first and come back

    line 43
  17. Very specific:

    Do NOT ask for or set ANTHROPIC_API_KEY

    line 45
  18. Very specific:

    Check if ANTHROPIC_API_KEY is already set in .env

    line 48
  19. Very specific:

    If not, ask for their key and append to .env

    line 49
  20. Distinctive:

    Verify it completes with "Migrations complete."

    line 56
  21. Distinctive:

    Use AskUserQuestion for each selection below.

    line 60
  22. Distinctive:

    Use AskUserQuestion (single select) with question "What kind of assistant do you want?":

    line 64
  23. Distinctive:

    Use AskUserQuestion (multi-select) with question "Which capabilities should be enabled?":

    line 80
  24. Very specific:

    All should be selected by default.

    line 86
  25. Distinctive:

    Use AskUserQuestion (multi-select) with question "Which integrations do you want enabled?":

    line 90
  26. Very specific:

    If the user selects GitHub or Sentry but the corresponding token is not set in .env, tell them after saving the config:

    line 97

“Common” is measured: kept by at least 6 in 10 of the 169 edited versions of this file we could read. “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

/setup — IronClaw Setup

Walk the user through setting up IronClaw from scratch.

IMPORTANT: Use the AskUserQuestion tool for ALL user choices. This renders clickable options instead of requiring the user to type. Never ask the user to "pick a number" or type a selection — always use AskUserQuestion with appropriate options.

Step 1: Check Prerequisites

  • Check Node.js version: run node --version
  • If Node.js is missing or below v22:
    • Try nvm install (which reads .nvmrc) then nvm use
    • If nvm is not available, tell the user: "Node.js 22+ is required. Please install it (e.g., using Node Version Manager with nvm install) or from https://nodejs.org, then re-run /setup."
    • Stop setup if Node.js 22+ cannot be activated
  • Verify pnpm is available (pnpm --version). If missing, run corepack enable && corepack prepare pnpm@latest --activate
  • Verify Docker is running (docker info)
  • If Docker is not running, tell the user to start Docker Desktop and re-run /setup

Step 2: Install Dependencies

  • Run pnpm install from the project root
  • Verify it completes without errors

Step 3: Start Infrastructure

  • Check if Docker Compose services are already running: docker compose -f docker-compose.dev.yml ps
  • If not running, start them: docker compose -f docker-compose.dev.yml up -d
  • Wait for Postgres to be healthy
  • Enable pgvector: docker exec ironclaw-postgres-1 psql -U ironclaw -d ironclaw -c "CREATE EXTENSION IF NOT EXISTS vector;"

Step 4: Authentication

Use AskUserQuestion (single select) with:

  • Claude Code Max plan — "Use your existing Claude subscription credits"
  • Anthropic API key — "Use a separate API key"

If user picks Max plan:

  • Run claude auth status to verify they're authenticated
  • If not authenticated, tell them to run claude auth login first and come back
  • Append AUTH_MODE=max_plan to .env (create if needed)
  • Do NOT ask for or set ANTHROPIC_API_KEY

If user picks API key:

  • Check if ANTHROPIC_API_KEY is already set in .env
  • If not, ask for their key and append to .env
  • Append AUTH_MODE=api_key to .env
  • Remind the user that .env is in .gitignore

Step 5: Run Database Migrations

  • Run pnpm db:migrate
  • Verify it completes with "Migrations complete."

Step 6: Configure Your Agent

Use AskUserQuestion for each selection below.

Persona

Use AskUserQuestion (single select) with question "What kind of assistant do you want?":

  • General assistant — "Good at everything (default)"
  • Developer assistant — "Code, debugging, architecture"
  • Research assistant — "Thorough research, citations, analysis"
  • Personal organizer — "Tasks, scheduling, reminders"

Note: AskUserQuestion supports max 4 options. Present the first 4, then if the user selects "Other", present a second question with the remaining options:

  • Writing assistant — "Content creation, editing, copywriting"
  • Data analyst — "Data exploration, SQL, visualizations"
  • DevOps / SRE — "Infrastructure, deployment, monitoring"
  • Product manager — "Requirements, specs, user stories"

The user can also type a custom description via the "Other" option on either question. If they do, set persona to "custom" and store their text in customPersona.

Boundaries

Use AskUserQuestion (multi-select) with question "Which capabilities should be enabled?":

  • Run bash commands — "Execute shell commands"
  • Write/edit files — "Create and modify files"
  • Web search — "Search the internet for information"
  • Read system files — "Access files outside the project directory"

All should be selected by default. Items the user deselects become false in boundaries.

MCP Connections

Use AskUserQuestion (multi-select) with question "Which integrations do you want enabled?":

  • Memory (built-in) — "Persistent long-term memory (recommended)"
  • GitHub (set GITHUB_TOKEN) — "Read issues, open PRs, search repos. Requires a GitHub personal access token in .env."
  • Sentry (set SENTRY_AUTH_TOKEN) — "List issues, inspect events. Requires a Sentry auth token in .env."

The memory MCP is enabled in-process via mcpConnections. GitHub and Sentry MCPs are gated by environment variables, not by mcpConnections — selecting them here is informational, the actual enablement happens when the corresponding token is present in .env at runtime.

If the user selects GitHub or Sentry but the corresponding token is not set in .env, tell them after saving the config:

"GitHub/Sentry will activate once you add GITHUB_TOKEN=ghp_… (or SENTRY_AUTH_TOKEN=sntrys_…) to your .env and restart IronClaw. See the README's Adding integrations section for the full walkthrough."

Save Configuration

Save the config by running from the shared package directory:

cd packages/shared && node --env-file=../../.env --import tsx/esm -e "
import { agentRepo } from './src/index.js';
const tenant = await agentRepo.findOrCreateDefaultTenant();
const agent = await agentRepo.findOrCreateDefaultAgent(tenant.id);
await agentRepo.updateAgentConfig(agent.id, {
  persona: 'CHOSEN_PERSONA',
  customPersona: 'CUSTOM_TEXT_OR_UNDEFINED',
  boundaries: {
    allowBash: BOOL,
    allowFileWrites: BOOL,
    allowWebSearch: BOOL,
    allowSystemFiles: BOOL
  },
  mcpConnections: ['memory']
});
console.log('Agent configured!');
process.exit(0);
"

Replace the placeholder values with the user's actual choices. Omit customPersona if persona is not "custom".

Step 7: Verify

Show a summary:

  • Auth mode: Max plan / API key
  • Persona: (chosen persona)
  • Boundaries: list enabled/disabled
  • MCP: list active connections

Then show how to start:

  • CLI mode: pnpm start
  • Full stack (one terminal): pnpm start:all
  • Full stack (separate terminals): pnpm start:gateway, pnpm start:runtime, pnpm start:web

Mention they can run /configure anytime to change agent settings.

Notes

  • Postgres runs on port 5433 (not 5432) to avoid conflicts
  • NATS runs on port 4222 with JetStream enabled
  • The default model is claude-sonnet-4-20250514 — override with ANTHROPIC_MODEL env var
  • Web app runs on port 3000 at http://localhost:3000

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.