Invite-only access

Stop writing ideas.
Start evolving them.

Ideas don't fail because they're bad. They fail because they never evolve. ForkIdeas gives ideas the infrastructure they deserve — versioning, structured critique, and validation tracking.

Request InviteEnter with Invite

$ forkideas init "AI-powered onboarding for SaaS"

✓ Idea repo created — v0.1.0 (seed)

$ forkideas critique --depth=deep

⚡ 3 risks identified, 2 assumptions unvalidated

$ forkideas evolve

→ Committed v0.2.0 — thesis refined, ICP narrowed

The problem

Ideas deserve infrastructure.
They get a text file.

Ideas decay

Written once in a note, never revisited. No structure, no evolution, no outcome.

No feedback loop

No systematic way to challenge assumptions, find risks, or validate hypotheses.

AI is disposable

Chat-based AI produces one-off answers. Nothing persists, nothing compounds.

How it works

Three steps to structured evolution

01

Create Idea Repo

Define your thesis, assumptions, ICP, and problem. Structure from day one.

02

Get AI Critique

AI generates structured artifacts — risks, gaps, validation plans. Not chat. Artifacts.

03

Evolve with Versions

Accept or reject critiques. Each decision creates a new version. Ideas evolve, not change.

Idea Repo Structure

Like a GitHub README.
But for your idea.

IDEA.repov0.2.0
# core_thesis
statement: "AI-powered onboarding reduces SaaS churn by 40%"
icp: "B2B SaaS, 10-50 employees, product-led growth"
problem: "Users churn in first 7 days due to poor onboarding"
# assumptions
- [open]  "Users prefer guided tours over documentation"
- [validated] "Onboarding completion correlates with retention"
- [open]  "AI personalization increases engagement > 20%"
# evidence
- interview: "8/10 users skip docs, want interactive setup"
- market_data: "Competitors with guided onboarding: 2x retention"
# critique
- [high] "AI personalization requires significant training data"
- [medium] "Privacy concerns with behavioral tracking"
# validation_log
- 2025-03-15: assumption_1 → validated (A/B test, n=340)
- 2025-03-22: assumption_3 → inconclusive (need more data)
Versioning

Ideas don't change.
They evolve.

Every decision creates a new version. Every critique, validation, and pivot is tracked. Full history. Full accountability.

v0.1.0seed

Initial thesis. Broad ICP. Unvalidated assumptions.

v0.2.0framed

ICP narrowed after 8 interviews. 2 risks accepted. Thesis refined.

v0.3.0validated

Core assumption validated via A/B test. MVP scope defined.

AI Layer

AI produces artifacts.
Not chat messages.

AI reads your entire repo context and generates structured, reviewable artifacts. Accept, reject, or defer — every decision is versioned.

critique_report

Structured critique with severity levels and recommended follow-ups

risk_analysis

Identified risks with mechanisms and mitigation directions

validation_plan

Experiments mapped to assumptions with success/failure signals

mvp_spec

Scoped MVP with must-haves, deferrals, and assumptions being tested

pending
accepted → creates new version
rejected → logged, no version change

We are not for everyone.

ForkIdeas is invite-only. Access is limited to ensure quality of the community and depth of engagement. Each user gets a limited number of invite codes.

01

Request an invite or get a code from a member

02

Enter your invite code to unlock registration

03

Earn invite codes to bring others in

Ready to evolve your ideas?

Stop treating ideas as static notes. Give them the infrastructure to become validated solutions.

Request Invite