venture-bootstrap
Skill4 starsChanged 8 months agoMITView on GitHub
Create a private GitHub repo for a venture, scaffold it with custom agents, skills, docs, and push
- Reads credentials
- Commits and pushes
In this repository14 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- brand-setupSkill
- daily-digestSkill
- deploySkill
- ecosystem-scoutSkill
- process-approvalsSkill
- publish-contentSkill
- qa-checkSkill
- send-outreachSkill
- submit-for-approvalSkill
- venture-health-checkSkill
- venture-intakeSkill
- venture-staffingSkill
- weekly-reviewSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
22 rules. About 0 are common, 8 are distinctive, and 14 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
Distinctive · estimate
This author's own choices. Worth a look.
14
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 19
Create a private GitHub repo for a venture and scaffold it with everything it needs to operate.
- Very specific: line 27
Read portfolio/ideas/{slug}.md for the idea brief
- Very specific: line 28
Read portfolio/ideas/{slug}-team.md for the team design
- Very specific: line 29
Read portfolio/guardrails.md for operating rules
- Very specific: line 53
Read templates/venture-claude.md and customize with venture-specific content:
- Very specific: line 64
Read templates/venture-readme.md and customize:
- Very specific: line 76
Leave {{PLACEHOLDERS}} for sections that need more research
- Very specific: line 92
Read the appropriate template from templates/agents/{agent}.md
- Distinctive: line 94
Add venture-specific context, principles, and instructions
- Very specific: line 99
If the team design (portfolio/ideas/{slug}-team.md) includes an "Ecosystem Tools" table:
- Distinctive: line 102
Add integration notes as comments in the relevant agent's file
- Very specific: line 103
If the tool is an MCP server, add it to the agent's tools: list
- Very specific: line 108
Add integration guidance to docs/architecture.md
- Very specific: line 111
Document in plans/current.md under a "Future Enhancements" section
- Distinctive: line 125
Run the brand-setup skill to research domains, validate the venture name, and generate a brand kit:
- Distinctive: line 134
Check social handle availability (Twitter/X, LinkedIn, GitHub)
- Distinctive: line 195
If the venture is at VALIDATION or later stage and has deployable code (e.g., a landing page):
- Very specific: line 197
Check portfolio/guardrails.md Stage → Capabilities Matrix to confirm deployment is allowed
- Very specific: line 198
Run the deploy skill to deploy to Vercel free tier
- Distinctive: line 200
Update the Notion Ventures DB entry with the live URL
- Very specific: line 202
If the venture needs a database, set up Supabase via the deploy skill.
- Distinctive: line 232
Search for the "Ventures" database in Notion and create a new entry:
≈ “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.
Files
Venture Bootstrap — Create and Scaffold a Venture Repo
Create a private GitHub repo for a venture and scaffold it with everything it needs to operate.
Input
Venture slug: $ARGUMENTS
Step 1: Read Context
- Read
portfolio/ideas/{slug}.mdfor the idea brief - Read
portfolio/ideas/{slug}-team.mdfor the team design - Read
portfolio/guardrails.mdfor operating rules - Determine the GitHub username:
gh api user --jq .login
Step 2: Create GitHub Repo
gh repo create {slug} --private --description "{one-liner from idea brief}"Clone it locally:
cd $PROJECTS_DIR
gh repo clone {slug}
cd {slug}Step 3: Create Directory Structure
mkdir -p agents skills commands src docs marketing data plans/archive decisions metrics learnings scheduled qa/reportsStep 4: Scaffold CLAUDE.md
Read templates/venture-claude.md and customize with venture-specific content:
- Replace all
{{PLACEHOLDERS}}with actual values from the idea brief - Set venture name, slug, type, stage (VALIDATION), one-liner, vision, target users
- Fill in the agent team table
- Set current focus based on stage (VALIDATION → "Validate demand and refine positioning")
- Set initial success metrics
Write to {slug}/CLAUDE.md.
Step 5: Scaffold README.md
Read templates/venture-readme.md and customize:
- Replace all
{{PLACEHOLDERS}}with actual values - Include current status, team, metrics
- Link to all docs
Write to {slug}/README.md.
Step 6: Scaffold Documentation
For each doc template in templates/docs/:
- Read the template
- Customize with venture-specific content where possible (from the idea brief and team design)
- Leave
{{PLACEHOLDERS}}for sections that need more research
Create these files:
docs/architecture.md— Populate what we know (type-appropriate stack recommendations)docs/requirements.md— Populate from idea brief (problem, target users, initial user stories)docs/market-research.md— Populate from researcher's data in the idea briefdocs/business-model.md— Populate from strategist's model recommendationdocs/changelog.md— Initial entry: "Venture bootstrapped"
Also create:
docs/user-personas.md— From idea brief's customer segmentsdocs/gtm-strategy.md— Initial placeholder with venture type-appropriate structure
Step 7: Scaffold Agents
For each agent in the team design:
- Read the appropriate template from
templates/agents/{agent}.md - Apply the customizations specified in the team design
- Add venture-specific context, principles, and instructions
- Write to
{slug}/agents/{agent}.md
Step 7b: Apply Ecosystem Enhancements
If the team design (portfolio/ideas/{slug}-team.md) includes an "Ecosystem Tools" table:
- For each tool classified as Agent Enhancement:
- Add integration notes as comments in the relevant agent's file
- If the tool is an MCP server, add it to the agent's
tools:list - Reference the tool in the agent's instructions where applicable
- For each tool classified as Skill Component:
- Note the tool in the relevant skill's instructions
- Add integration guidance to
docs/architecture.md
- For each tool classified as Future-Stage Addition:
- Document in
plans/current.mdunder a "Future Enhancements" section - Note which stage to revisit
- Document in
- Reference all ecosystem tools in
docs/architecture.mdunder a "Ecosystem Integrations" section.
Step 8: Create Initial Skills
Based on the team design's skill list, create venture-specific skills:
At minimum, create:
skills/status-report/SKILL.md— Generate a status report for ChiefOfStaff health checks
Step 9: Brand Setup (if stage >= VALIDATION)
Run the brand-setup skill to research domains, validate the venture name, and generate a brand kit:
/brand-setup {slug}This will:
- Validate the venture name (trademark conflicts, SEO, pronunciation)
- Research domain availability (.com, .ai, .io, .co + variations)
- Check social handle availability (Twitter/X, LinkedIn, GitHub)
- Generate a complete brand kit at
docs/brand-kit.md(identity, voice & tone, visual identity, domain recommendations)
Skip this step if the venture is at IDEA stage (brand research happens during VALIDATION).
Step 10: Create Continuity File
Read templates/continuity.md and customize:
- Replace
{{VENTURE_NAME}}with the venture name - Set Last Updated to today's date, by "chief-of-staff"
- Set "What Was Done" to "Venture bootstrapped with agent team and docs"
- Set "What's Next" to the first task from the plan (typically: deep market research)
Write to {slug}/continuity.md.
Step 11: Create Initial Plan
Write plans/current.md:
# Current Plan — {Venture Name}
**Stage:** VALIDATION
**Sprint:** Initial Setup
**Updated:** {today}
## Objectives
1. Validate demand for {venture concept}
2. Identify and deeply understand target customer segment
3. Sharpen competitive positioning
4. Design first validation experiment
## Tasks
- [ ] Deep market research (researcher)
- [ ] Customer persona development (researcher)
- [ ] Competitive positioning document (researcher + marketer)
- [ ] Validation experiment design (marketer)
- [ ] Landing page or signup test plan (marketer)
## Success Criteria
- Market research completed with TAM/SAM/SOM
- At least 2 customer personas defined
- Clear differentiation from top 3 competitors
- One validation experiment designed and ready to executeStep 12: Create .gitignore
.DS_Store
node_modules/
.env
.env.*
*.log
dist/
build/
.claude/Step 13: Deploy (if stage allows)
If the venture is at VALIDATION or later stage and has deployable code (e.g., a landing page):
- Check
portfolio/guardrails.mdStage → Capabilities Matrix to confirm deployment is allowed - Run the
deployskill to deploy to Vercel free tier - Record the deployment URL in the venture's CLAUDE.md Deployment section
- Update the Notion Ventures DB entry with the live URL
If the venture needs a database, set up Supabase via the deploy skill.
Skip this step if:
- The venture is at IDEA stage (no deployment)
- There's no deployable code yet (just docs/plans)
Step 14: Initial Commit and Push
cd $PROJECTS_DIR/{slug}
git add -A
git commit -m "Bootstrap venture: {venture name}
Scaffolded by ChiefOfStaff with custom agent team.
Stage: VALIDATION
Type: {venture type}
Team: {agent list}"
git push -u origin mainStep 15: Update Portfolio Registry
Append the new venture to portfolio/registry.md:
| {slug} | {Venture Name} | github.com/{user}/{slug} | VALIDATION | -- | $0 | {today} | Begin validation |Step 16: Update Notion
Search for the "Ventures" database in Notion and create a new entry:
- Name: {Venture Name}
- Stage: VALIDATION
- Health: --
- MRR: $0
- Repo URL: https://github.com/{user}/{slug}
- Last Activity: {today}
- Next Action: Begin validation research
Step 17: Summary
Report to the founder:
Venture "{Venture Name}" bootstrapped successfully.
- GitHub repo: https://github.com/{user}/{slug} (private)
- Stage: VALIDATION
- Agent team: {list of agents}
- Docs: architecture, requirements, market-research, business-model, changelog
- Next: Run /run-cycle or wait for morning cycle to begin validation work
The agents are ready. Run `/venture-status {slug}` to check progress at any time.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.

