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
- NeverSight/CTO-SKILLStheir own version, first seen Oct 15, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- NeverSight/CTO-SKILLS★ 8same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository36 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude/AGENTS.mdAGENTS.md
- template/AGENTS.mdAGENTS.md
- .claude/CLAUDE.mdCLAUDE.md
- .claude/upstream/CLAUDE.mdCLAUDE.md
- agent-browserSkill
- algorithmic-artSkill
- api-testerSkill
- brand-guidelinesSkill
- canvas-designSkill
- code-reviewer-123Skill
- design-auditSkill
- design-brainSkill
- doc-coauthoringSkill
- docxSkill
- frontend-designSkill
- internal-commsSkill
- manus-kickoffSkill
- mcp-builderSkill
- my-skillSkill
- pdfSkill
- pptxSkill
- quality-guardSkill
- quick-fixSkill
- vercel-react-best-practicesSkill
- seo-optimizationSkill
- skill-creatorSkill
- slack-gif-creatorSkill
- template-skillSkill
- test-skill-nameSkill
- theme-factorySkill
- threejs-fundamentalsSkill
- webapp-testingSkill
- web-artifacts-builderSkill
- web-design-guidelinesSkill
- xlsxSkill
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
- Distinctive: line 15
Do not use it for small bug fixes or tiny scripts unless the user explicitly requests the cofounder workflow.
- Distinctive: line 22
Assumptions checklist first: before building, list assumptions.
- Very specific: line 24
Read PHASE=1..5 (or "Phase 1..5") from user input.
- Very specific: line 25
If missing, default to PHASE=1 (Discovery).
- Distinctive: line 26
Only progress one phase per run.
- Distinctive: line 50
Always output roles in order: Conductor → Architect → Builder → Reviewer.
- Distinctive: line 54
Architect/Builder/Reviewer: should output only placeholders unless the current phase explicitly requires them.
- Distinctive: line 69
Stop + confirm: ask PO to choose next step.
- Distinctive: line 99
V1 scope: must-have vs later, with explicit non-goals.
- Distinctive: line 124
Stop at decision points (A-G) if new info appears.
- Distinctive: line 129
No hidden “magic”; setup is documented.
- Very specific: line 133
Add/adjust a small scope item (still PHASE=3)
- Very specific: line 134
Stop and re-plan (back to PHASE=2)
≈ “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.
Files
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.
- Read
- 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:
- 关键问题(5-8 个): highest information gain, must cover A-G by combining questions when possible.
- 我将做的假设清单(可逐条否决): explicit checkboxes.
- V1 初步边界: “must-have now” vs “add later” (rough, reversible).
- 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):
- Provide answers + confirm assumptions → proceed to
PHASE=2 - Adjust scope/assumptions first (stay in
PHASE=1) - 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:
- V1 scope: must-have vs later, with explicit non-goals.
- Technical approach (plain language): architecture sketch, data flow, key modules.
- Accounts/services needed: list with ownership and links/notes (no secrets).
- Rough outline of finished product: screens, flows, APIs, background jobs (as applicable).
- 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:
- Approve and proceed to
PHASE=3 - Revise V1 scope/architecture (stay in
PHASE=2) - 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:
- Proceed to
PHASE=4(Polish) - Add/adjust a small scope item (still
PHASE=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:
- Proceed to
PHASE=5(Handoff) - 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.
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
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.

