start

An interactive setup wizard for Claude Code, a coding assistant that works in your project. It examines the codebase, asks about your project and workflow, then prepares tailored project tooling.

In plain words
What is it for?
Use it to onboard an existing project, research its structure and risks, detect installed plugins, and preview a generated Claude Code setup. It runs only when explicitly started with `/onboard:start`.
Why use it?
It avoids starting with a generic setup that may not fit the project. The guided review also makes its findings and proposed changes visible before approval.

Skill for Claude CodeCodex

Part of the onboard plugin — 8 skills, 5 agents shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/apurvbazari/claude-plugins/start
Any agent
npx skills add ApurvBazari/claude-plugins --skill start
Clone the repo
git clone --depth 1 https://github.com/ApurvBazari/claude-plugins

Made for: Claude Code, Codex.

Or install onboard, the plugin that ships this one along with the rest of its 8 skills, 5 agents.

Per session 35 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 7,488 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00035 $0.07488
Opus 5 $0.00017 $0.03744
Sonnet 5 $0.00007 $0.01498
Haiku 4.5 $0.00003 $0.00749

Measured 2d ago against content hash f7bb77ff4506, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

start scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

onboard/skills/start/SKILL.md · 404 lines

How it starts

The opening of the file, as written. The whole thing — 404 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Start Skill — Interactive Onboarding Wizard

You are running the onboard start skill. This is a guided, multi-phase process that analyzes a developer's codebase and generates complete Claude tooling infrastructure.

Overview

Tell the developer:

Starting onboard — I'll analyze your codebase, walk you through some questions about your project and workflow, then generate a complete Claude Code setup tailored to your project.

This runs in these phases:

  • Phase 0 — Empty-repo guard: if there's no source code yet, I route to a minimal stub instead of the full flow
  • Phase 1 — Recon: I scan your codebase (read-only, native tools)
  • Phase 2 — Research: you pick a depth/scope profile (Minimal / Standard / Comprehensive), then focused specialists investigate per dimension and I verify their findings, writing a research dossier + architecture map + risk register + glossary
  • Phase 3 — Grounded Wizard: I show you what research inferred; you confirm or override
  • Phase 4 — Plugin detection & context: I detect installed plugins and assemble the generation context
  • Phase 5 — Plan → Preview → gate: I show a full preview behind a hard Approve / Adjust / Cancel gate; nothing is written until you approve
  • Phase 6 — Generation: I create all Claude tooling artifacts
  • Phase 7 — Handoff: I explain everything that was generated

Phase 0: Empty-Repo Guard

Runs before Phase 1 Recon. Detects repositories with no source code and routes them to a minimal, canonical-shape stub instead of running the full analysis + wizard. Closes 2026-04-17 release-gate findings B14, B15, B16.

Follow references/empty-repo-stub-procedure.md in full — it is the single source of truth for the detection filter (SRC_COUNT), the prior-stub re-run check, the 3-option AskUserQuestion menu (Abort / Placeholder only / Generate canonical stub), and each option's execute path.

A stub, abort, or placeholder run returns from that procedure directly — before Step 0 below — and never reaches Phase 1 Recon, so no phase-task list is created for it.

Read the full file on GitHub · 404 lines

Files

What ships with it

5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 404 lines · 35 tokens per session scan A f7bb77ff4506

Subscribe to this mod's changes

start is a skill published in the GitHub repository ApurvBazari/claude-plugins (0 stars, last pushed 26d ago), licensed MIT. It adds 35 tokens to every session and 7,488 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.