Skip to content
Get packs

Before You Build

Runs a quick pre-mortem on a product or feature idea before any code is written

Source Download skill
Repository
wshobson/agents
Stars
42,095
Licence
MIT
Files
2 files

What it does

When you ask your AI agent to build a landing page, MVP, app, or major feature, this skill makes it pause and check the idea first. It looks at demand, positioning, money, retention, trust, distribution and adoption risk, then names the single riskiest assumption, the cheapest way to test it, and what to delay building. It is a thinking aid written as instructions, not a tool that goes out and gathers market data.

Great for

  • Founders or product managers about to commit engineering time to a new product or feature
  • AI-assisted builders who tend to start coding before checking whether anyone wants the thing
  • Teams planning a public launch who want to test positioning before spending on promotion

Skip it if

  • Narrow implementation work such as bug fixes, refactors, test repairs or dependency updates, which the skill itself says to skip; it adds delay without value there
  • Making a final go/no-go decision: the output is a judgement based only on what you tell the agent, with no real market evidence checked, so it can be wrong about your market

Try asking

  • “Before I build an MVP for a freelance invoicing app, run the before-you-build review”
  • “Check the demand and monetization risks for an internal agent workflow we want to build this week”
  • “Review my landing page idea for a meal-planning SaaS before I start coding it”
Step-by-step guide: how to use Before You BuildInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 2 original files, the MIT licence and an attribution note.

    before-you-build.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip before-you-build.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip before-you-build.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Describe the target user, the problem and any evidence you already have (interviews, sign-ups, waitlist numbers) so the review has something concrete to assess
  • Ask for the smallest validation step rather than a full research plan, since the skill is designed for compact answers
  • Treat the verdict as a prompt for your own customer conversations, not a substitute for them

What's inside

  • references/risk-checklist.md
  • SKILL.md

Redistributed unmodified from wshobson/agents under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.