agentleFS
Sign inStart free
patmeh1patmeh1/product-docs-kit/.claude/skills/bootstrap-docs/SKILL.md

bootstrap-docs

Skill0 starsChanged 2 months agoMITView on GitHub

Generate a complete product documentation set from a product idea when little or no code exists yet. Use when the user has an idea and a filled-in vision document and wants a PRD, user flows, design system, architecture, data model, API contract, tech stack, security model, and test plan created from scratch. Also use for "start a new project", "plan my app", or "turn my idea into a spec".

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

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

    Read docs/00-vision.md first: If it is still full of _TBD_, stop and tell the user to fill it in by hand.

    line 12
  2. Very specific:

    Read the full procedure in prompts/01-greenfield-bootstrap.md and follow it exactly.

    line 16
  3. Distinctive:

    Ask before assuming: If the vision document does not answer something that materially changes the design, ask rather than picking for the …

    line 21
  4. Very specific:

    Be strict on priority: If more than roughly 60% of requirements are Must, the user has not prioritised — rework it.

    line 44

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Greenfield Bootstrap — idea to full specification

Take a product idea to a documented, buildable specification.

Before you start

  1. Read docs/00-vision.md first. If it is still full of _TBD_, stop and tell the user to fill it in by hand. You must not invent business intent, and everything downstream derives from it.
  2. Read AGENTS.md.
  3. Read the full procedure in prompts/01-greenfield-bootstrap.md and follow it exactly.

Rules

  • Ask before assuming. If the vision document does not answer something that materially changes the design, ask rather than picking for the user.
  • Mark every choice you made with [ASSUMPTION], and state the alternative you rejected.
  • Scope discipline. Everything you propose must trace to an OBJ-. If it does not, leave it out and list it separately as a deferred idea.
  • Prefer boring technology. Well-understood tools unless there is a stated reason otherwise. Novel choices cost debugging time the user has not budgeted.
  • Design for v0.1 — the walking skeleton — not for imagined v3 scale.

Stages — stop for the user after each

StageProduces
0Your understanding, gaps in the vision, the riskiest assumption, the smallest viable v0.1, and clarifying questions. No documents yet.
101-prd.md — then present the requirements as a table and confirm scope and priority
202-user-flows.md, 03-design-system.md
304-architecture.md, 05-data-model.md, 06-api-contract.md, 12-tech-stack.md, plus ADRs
407-security-privacy.md, 08-test-plan.md, 10-release-runbook.md, 11-glossary.md
509-traceability-matrix.md, a build order table, AGENTS.md, CHANGELOG.md

Specifics worth getting right

Be strict on priority. If more than roughly 60% of requirements are Must, the user has not prioritised — rework it.

Pin versions in 12-tech-stack.md, not "latest". Record each technology's end-of-life horizon so upgrade planning starts on day one. Prefer long-term-support releases.

Verify contrast ratios for every colour pairing you propose in the design system, and state them.

Sequence the build order to get a walking skeleton — one complete end-to-end path through the real stack — working as early as possible.

On finishing

Set every document to version: 0.1.0, status: Draft, and add the first Change Log row. Do not mark anything Approved.

Summarise every [ASSUMPTION] you made, every open question, and the three biggest risks in the plan.

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.

Check your setup, free