Prompt pack · Vibe coding

Spec-First Build Docs: PRD, design system, architecture note, agents guide and readiness check

by PaperChaseWebb

A Markdown pack of the four documents to write before an agent builds anything, sized to the stakes: a PRD an agent can build from without inventing scope, the smallest design system that holds, a one-page architecture note with reversal costs, an agents guide with a project tree for Claude Code, Codex and Cursor, a weekly staleness pass and a nine-area readiness check.

Preview card for Spec-First Build Docs
Works withClaude CodeCodexCursorPlain .md
Price
$29
Difficulty
Intermediate
Setup
30 min
Format
.md file
$29 one time

spec-first-build-docs-prd-design-system-architecture-agents-guide.md · 26 KB · instant download after checkout

Buy now · $29

Want this built for you instead? We can build it for you.

What you end up with

Your coding agent carries out decisions you wrote down instead of guessing, the docs stay true as the project changes, and you know what is missing before real users arrive.

What you get

  • 10 sections in one .md file
  • Three spec weights (one page, standard, full) and a one-minute way to pick
  • A PRD template with a filled example, failure steps and a Decisions owed table
  • A small design system template with color roles, component states and a never list
  • A one-page architecture note with boundaries, failure behavior and a reversal cost table
  • An agents guide template and a project tree for Claude Code, Codex and Cursor
  • A weekly staleness pass with update triggers
  • A nine-area production readiness check with a one-page verdict
  • 12 prompts to draft, check and maintain the set

Who it is for

Builders who start agent projects from a prompt and keep paying for the agent's guesses in rewrites.

Works with

Claude CodeCodexCursorPlain .md

Solid chips: the agent is named in this listing's own tools list. Dashed chip: Plain Markdown, so it also pastes into other agents such as Grok Bot, Hermes, OpenClaw or Muse. We list those as compatible by format only, not as tested.

Tools needed

Claude Codereads CLAUDE.md and the docs folder
Codexreads AGENTS.md
Cursorreads the project rule that points to the guide
You needA coding agent such as Claude Code, Codex or Cursor
You needA project folder or repo

Cost to run: Free. Uses the coding agent you already have.
Time to result: A working PRD and agents guide in one sitting

Sample prompts, free to use

Find the guesses · Claude Code, Codex, Cursor
Read this PRD [paste] as the agent that will build it.
List every place you would have to guess: a missing failure step, a feature that isn't testable, a data field with no sensitivity, a decision hidden in a sentence. For each, write the question I need to answer.

Run it before the first build prompt, not after.

Gather audit evidence · Claude Code, Codex
For this system, go through the nine readiness areas. For each check, point to the file, setting or log that answers it, or write NO EVIDENCE.
Do not mark anything Pass. A person will grade it.

The builder gathers evidence. Someone else signs the verdict.

Free preview of the file

spec-first-build-docs-prd-design-system-architecture-agents-guide.md (excerpt)26 KB total
# Spec-First Build Docs: PRD, design system, architecture note, agents guide and readiness check

A PaperChaseWebb pack, version 1.0. Written for Claude Code, Codex and Cursor, and plain Markdown, so every template also works with any agent that reads files from your repo.

A coding agent builds from whatever context it has. When the scope, the look, the stack and the rules live only in your head, the agent fills the gaps with guesses, and you pay for those guesses later in rewrites. This pack gives you the four documents to write before the first prompt, a project tree that keeps them in front of the agent, a pass that keeps them true as the project changes, and an audit to run before real users arrive. It is sized to the stakes: a weekend experiment gets one page, a client build gets the full set.

It comes from how we start our own agent builds, plus three public breakdowns we studied and rebuilt in our own words. Every source is credited by link in `09-sources.md`.

## Files

| File | What it is for |
|---|---|
| 01-size-the-gate.md | Three weights of spec (one page, standard, full) and how to pick one in a minute |
| 02-prd.md | A product requirements template an agent can build from without inventing scope, with a filled example |
| 03-design-system.md | The smallest design system that keeps agent-built screens consistent: tokens, components, states, banned moves |
| 04-architecture-note.md | One page that fixes the expensive decisions: stack, data flow, boundaries, failure behavior, reversal cost |
| 05-agents-guide.md | The rules file that governs every agent in the repo, plus a project tree for Claude Code, Codex and Cursor |
| 06-staleness-pass.md | A short routine that keeps the four documents true, with triggers that force an update |
| 07-readiness-audit.md | A nine-area production readiness check graded from evidence, with a one-page verdict |
| 08-prompts.md | 12 prompts to draft, check and maintain the whole set |
| 09-sources.md | Every source we used, linked and credited |

Worth knowing

  • Tool file names and settings change often. Check the current Claude Code, Codex and Cursor docs for where instruction files live.
  • The readiness check is a starting checklist, not a certification. Systems handling payments, health or personal data need the standards that apply to their field.

Questions

Do I need all four documents for a weekend project?

No. The pack sizes the gate to the stakes. A spike gets a one-page spec that answers the same four questions in a few lines each.

How is this different from the AGENTS.md and CLAUDE.md starter files?

The starter files cover the rules file alone. This pack covers what the rules file points to: the PRD, design system and architecture note, how to keep them current, and an audit before launch.

Details

Version 1.0 · Updated Oct 26, 2026 · Markdown version of this page
License: Personal and commercial use. Your specs are yours. Do not resell or republish this pack. Download terms

specsprdagent-rulesclaude-codecodexcursorproduction-readinesstemplates

Related

Preview card for AGENTS.md, CLAUDE.md and Cursor rules starter files
Prompt pack$14

AGENTS.md, CLAUDE.md and Cursor rules starter files

by PaperChaseWebb

Six fill-in instruction files for coding agents (base, Next.js plus Supabase, FastAPI, browser game, CLAUDE.md pointer, scoped Cursor rules) and 8 prompts to write, trim and test your own.

Works withClaude CodeCodexCursorPlain .md
Vibe coding · Beginner · 20 min setup
Preview card for Pre-launch security prompts for vibe-coded apps
Prompt pack$19

Pre-launch security prompts for vibe-coded apps: Supabase RLS, leaked keys and open doors

by PaperChaseWebb

14 prompts that make your coding agent find exposed keys, tables without row level security, open buckets and unverified webhooks in your own app, then fix them one commit at a time.

Works withClaude CodeCodexCursorWindsurfPlain .md
Vibe coding · Beginner · 15 min setup
Preview card for App Launch Check
App reviewFrom $29

App Launch Check: a launch review for apps built with AI

by PaperChaseWebb

Send a repo URL and get a scored report on deploy, domain, payments, legal pages, security basics and AI code mistakes, with a fix list in priority order.

Works withService, no agent needed
Vibe coding · Beginner · 5 min setup
Preview card for Vibe Coding with AI Agents
Prompt pack$39

Vibe Coding with AI Agents: method, prompts, skills and checklists

by PaperChaseWebb

A Markdown playbook for building real software with Claude Code, Codex or Claude: a seven-step method, project file templates, 18 prompts, five agent skills, release and hardware checklists, cost notes and 24 project ideas.

Works withClaude CodeCodexClaudePlain .md
Vibe coding · Intermediate · 20 min setup
Preview card for Governed Agent Teams
Prompt pack$39

Governed Agent Teams: subagent personas, authority maps and handoffs

by PaperChaseWebb

A Markdown pack for running a team of coding and research agents with clear limits: 7 subagent personas for Claude Code with Codex AGENTS.md and Hermes skill versions, a Desk Authority Map, three role risk classes, a handoff packet the receiver verifies, an evidence standard, a decision ledger and a reach inventory.

Works withClaude CodeCodexHermesPlain .md
Solo business · Intermediate · 20 min setup