README.md
A persistent startup copilot from problem discovery to first revenue. Tracks customer evidence, competitor intelligence, validation experiments, and strategic decisions — and pushes back when you try to build before you've validated.
This is not a note-taking tool. It is startup execution cognition.
Load the built plugin folder in LM Studio.
| Field | Default | Description |
|---|---|---|
| Data Path | ~/founder-data/ | JSON file location for all founder data |
| SearXNG URL | (blank) | Self-hosted SearXNG for competitor research. Falls back to DDG → Bing if blank |
| Stage | What it means |
|---|---|
idea | Hypothesis exists, no customer evidence yet |
problem | Customer problems logged from real interviews/observations |
solution | MVP scoped, first experiments running |
traction | First paying users, validating willingness to pay |
growth | Retention proven, scaling distribution |
paused / killed | Inactive — decision made to stop |
| Tool | Kind | What it does |
|---|---|---|
capture_startup | Store-write | Register a new startup with hypothesis, ICP, problem statement |
update_startup | Store-write | Update fields as the startup evolves — stage, pivot, ICP refinement |
list_startups | Store-read | List all startups with stage and one-liner |
get_startup | Store-read | Full detail for one startup including counts of evidence |
| Tool | Kind | What it does |
|---|---|---|
log_customer_problem | Store-write | Record a pain point with verbatim quote, source, frequency, severity |
list_customer_problems | Store-read | List problems sorted by severity and frequency |
| Tool | Kind | What it does |
|---|---|---|
scan_competitors | Effect + Store-write | Search for competitors and save positioning data |
update_competitor | Store-write | Add pricing, strengths, weaknesses, differentiator |
list_competitors | Store-read | List all tracked competitors for a startup |
| Tool | Kind | What it does |
|---|---|---|
create_validation_plan | Scaffold | Map assumptions, rank by risk, recommend cheapest experiments |
record_experiment_result | Store-write | Log what was tested, result, learnings, and evidence |
list_experiments | Store-read | List all experiments with results and learnings |
| Tool | Kind | What it does |
|---|---|---|
log_decision | Store-write | Record a strategic decision with options and rationale |
update_decision_outcome | Store-write | Record what actually happened as a result |
list_decisions | Store-read | Decision log in reverse chronological order |
| Tool | Kind | What it does |
|---|---|---|
generate_mvp_scope | Scaffold | Define the smallest testable version of the core hypothesis |
launch_asset_generator | Scaffold | Landing page, cold outreach, elevator pitch, positioning statement |
| Tool | Kind | What it does |
|---|---|---|
weekly_founder_review | Store-read | Cross-startup review — what's untested, pending, and most urgent |
| Type | What it means |
|---|---|
concierge | Manually deliver the service. No code. Validates willingness to pay before building. |
wizard_of_oz | Front-end only — you do the back-end manually. Users think it's automated. |
landing_page | One page with CTA. Measures demand before building anything. |
single_feature | One feature, nothing else. Tests the core value hypothesis only. |
prototype | Clickable Figma/Framer mock. Zero code. Tests UX and value prop. |
The plugin enforces evidence-based decisions:
The plugin will push back if you try to:
New startup:
"I want to build a tool that helps freelancers track unpaid invoices"
list_startups → capture_startup(name="InvoiceGuard", hypothesis="Freelancers lose money because they forget to follow up on late invoices", icp="Freelancers earning $50k–$150k/yr") → "Have you talked to any freelancers about this?"
After customer interviews:
"I talked to 5 freelancers — they all said their biggest problem is not knowing which clients are likely to ghost"
log_customer_problem(source="interview", description="Can't predict which clients will go silent", verbatim="I never know if I should follow up or if they've just gone dark", severity="critical", frequency="frequent")
Validation planning:
"What should we test first?"
create_validation_plan(startupId) → returns ranked assumptions with recommended experiments → cheapest first: landing page for demand, then concierge for willingness to pay
Weekly check-in:
"Weekly review — what should I focus on?"
weekly_founder_review() → surfaces pending experiments, open decisions, gaps in evidence, and the single most important action per startup
Writing the launch page:
"Help me write the landing page for InvoiceGuard"
launch_asset_generator(startupId, assetType="landing_page", ctaGoal="waitlist", tone="empathetic") → returns headline options, subheadline, feature bullets, CTA copy, FAQ
cd founder-plugin
npm install
npm run build
idea → problem → solution → traction → growth