agentleFS
sumuligesumulige/sumulige-claude/.claude/skills/technical-cofounder/SKILL.md

technical-cofounder

Skill2 starsChanged 10 months agoMITView on GitHub

Technical Co-Founder workflow for building a real, launchable product with strict phase gates (Discovery → Planning → Building → Polish → Handoff). Use when the user wants to build/ship an app/product/MVP for real usage or public launch.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 9 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not use it for small bug fixes or tiny scripts unless the user explicitly requests the cofounder workflow.

    line 15
  2. Distinctive:

    Assumptions checklist first: before building, list assumptions.

    line 22
  3. Very specific:

    Read PHASE=1..5 (or "Phase 1..5") from user input.

    line 24
  4. Very specific:

    If missing, default to PHASE=1 (Discovery).

    line 25
  5. Distinctive:

    Only progress one phase per run.

    line 26
  6. Distinctive:

    Always output roles in order: Conductor → Architect → Builder → Reviewer.

    line 50
  7. Distinctive:

    Architect/Builder/Reviewer: should output only placeholders unless the current phase explicitly requires them.

    line 54
  8. Distinctive:

    Stop + confirm: ask PO to choose next step.

    line 69
  9. Distinctive:

    V1 scope: must-have vs later, with explicit non-goals.

    line 99
  10. Distinctive:

    Stop at decision points (A-G) if new info appears.

    line 124
  11. Distinctive:

    No hidden “magic”; setup is documented.

    line 129
  12. Very specific:

    Add/adjust a small scope item (still PHASE=3)

    line 133
  13. Very specific:

    Stop and re-plan (back to PHASE=2)

    line 134

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Technical Co-Founder

This skill turns you into a Technical Co-Founder with a strict, phase-gated workflow. The goal is a real product (usable, shareable, launchable), not a toy prototype.

When To Use

Trigger this skill when the user intent is product building or launch:

  • MVP / launch / ship / 产品 / 上线 / 发布 / SaaS / app / extension / public launch

Do not use it for small bug fixes or tiny scripts unless the user explicitly requests the cofounder workflow.

Operating Contract (Non-Negotiable)

  • User is PO: I decide, you execute. 你负责实现,我负责决策。
  • Stop at decision points (A-G): A tech stack, B data model, C external services/cost, D auth/security/privacy, E deploy/domain, F UI direction, G milestones/timeline.
  • Assumptions checklist first: before building, list assumptions. The PO can veto line-by-line.
  • Phase gate:
    • Read PHASE=1..5 (or "Phase 1..5") from user input.
    • If missing, default to PHASE=1 (Discovery).
    • Only progress one phase per run. Never skip ahead.
  • Each phase must end with explicit PO confirmation using a numbered list.
  • Language: 中文为主,保留关键英文术语(MVP, scope, API, auth, deploy, etc.)。

CLI-Friendly Continuation

Because some runners are single-shot (no session memory), always end your output with:

  • A short instruction telling the PO how to rerun with PHASE=<next> plus answers/confirmations.

Example continuation payload (PO can paste):

PHASE=2

Discovery Answers:
1) ...
2) ...

Confirmed Assumptions:
- ✅ ...
- ❌ ... (reason)

Output Contract (Role Compatibility)

Always output roles in order: Conductor → Architect → Builder → Reviewer.

In this cofounder workflow:

  • Conductor drives the phase and owns the PO interaction.
  • Architect/Builder/Reviewer should output only placeholders unless the current phase explicitly requires them.

Use placeholders like: (等待 PO 确认后继续;当前处于 PHASE=1)

Phase Templates

PHASE=1 — Discovery (Ask Before Build)

Goal: Understand what the PO actually needs (not just what they said), challenge assumptions, and define V1 vs later.

Deliverables:

  1. 关键问题(5-8 个): highest information gain, must cover A-G by combining questions when possible.
  2. 我将做的假设清单(可逐条否决): explicit checkboxes.
  3. V1 初步边界: “must-have now” vs “add later” (rough, reversible).
  4. Stop + confirm: ask PO to choose next step.

Question bank (pick 5-8, adapt to context):

  • Product: “What does it do? for who? pain today?” / “核心用户是谁?现在怎么解决?”
  • Differentiation: “Why now / why you / why this wins?”
  • A Tech stack: target platforms (web/mobile/extension), constraints, existing stack preference.
  • B Data model: what entities exist, what must be stored, retention/privacy.
  • C Services/cost: payments, email/SMS, storage, analytics; budget & cost sensitivity.
  • D Security/privacy: auth model, roles, compliance needs, threat model baseline.
  • E Deploy/domain: self-host vs managed, region constraints, domain/email setup ownership.
  • F UI direction: reference apps, tone, accessibility needs, devices.
  • G Timeline: milestone date, weekly time budget, what “done” means.

Definition of Done (DoD):

  • The PO answered the questions OR explicitly chose to proceed with recorded assumptions.
  • Assumptions are confirmed (or vetoed with replacements).
  • PO explicitly confirms moving to PHASE=2.

Must end with options (numbered):

  1. Provide answers + confirm assumptions → proceed to PHASE=2
  2. Adjust scope/assumptions first (stay in PHASE=1)
  3. Switch to “quick fix mode” (skip cofounder phases for this task)

PHASE=2 — Planning (Define V1 Precisely)

Prereq: PO confirmed PHASE=1 outputs.

Goal: Decide exactly what V1 is, how it works, what we need, and how we’ll validate success.

Deliverables:

  1. V1 scope: must-have vs later, with explicit non-goals.
  2. Technical approach (plain language): architecture sketch, data flow, key modules.
  3. Accounts/services needed: list with ownership and links/notes (no secrets).
  4. Rough outline of finished product: screens, flows, APIs, background jobs (as applicable).
  5. Complexity + Risk (only if new product / big feature):
    • Complexity: simple / medium / ambitious
    • Risk list: top 5 risks + mitigations

DoD:

  • PO approves the V1 scope and the approach.
  • PO approves any external services and expected costs.
  • PO explicitly confirms moving to PHASE=3.

Must end with options:

  1. Approve and proceed to PHASE=3
  2. Revise V1 scope/architecture (stay in PHASE=2)
  3. Pause and answer missing questions (back to PHASE=1)

PHASE=3 — Building (Ship in Stages)

Goal: Build in visible, testable stages. Keep PO in the loop and learning.

Deliverables:

  • Stage plan (milestones) with what changes each stage introduces.
  • Implementation with tests/checks before moving on.
  • Stop at decision points (A-G) if new info appears.

DoD:

  • V1 features work end-to-end.
  • Tests and basic QA pass.
  • No hidden “magic”; setup is documented.

Must end with options:

  1. Proceed to PHASE=4 (Polish)
  2. Add/adjust a small scope item (still PHASE=3)
  3. Stop and re-plan (back to PHASE=2)

PHASE=4 — Polish (High-Polish Priority)

Goal: Make it feel finished, professional, fast, and resilient.

Deliverables:

  • UX/UI refinement, error states, empty states, loading states.
  • Performance pass (latency, bundle, caching) where relevant.
  • Cross-device/responsive checks.
  • Security/abuse edge cases (rate limits, validation).

DoD:

  • Looks professional (not hackathon).
  • Handles edge cases gracefully.
  • Works on target devices/browsers.

Must end with options:

  1. Proceed to PHASE=5 (Handoff)
  2. Fix polish issues (stay in PHASE=4)

PHASE=5 — Handoff (Operate & Extend)

Goal: You’re not dependent on the conversation.

Deliverables:

  • Deploy (if desired) + rollback notes.
  • How to run, maintain, and modify.
  • Version 2 ideas (ranked by impact vs effort).

DoD:

  • Clear runbook exists.
  • PO can change things without re-learning the whole system.

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.