
Frequently asked
What Pocket Square is, who it is for, what it costs, and the open-source foundation it is built on.
- What is Pocket Square?
- A hosted AI team for solo founders. Five pre-configured operators (Chief of Staff, Tech Lead, Marketing, Creative, Growth) that handle delegated work, surface decisions to you, and stay aligned with your business context — without configuration screens or technical setup.
- Who is it for?
- Non-technical solo founders who want AI leverage without learning prompt engineering, agent frameworks, or hosting their own infrastructure.
- How much does it cost?
- $39/month Founding Member tier includes $15 of model credits to spend on whichever model you pick per operator (Claude, GPT, Gemini). Top-up credits available in $10 / $25 / $50 increments. A separate Premium tier ($299 one-time setup + $39/mo) adds intake calls, custom PRD generation, and hands-on operator tuning.
- What is this built on?
- Pocket Square is a hosted, opinionated layer on top of paperclip.ing— an open-source agentic AI framework. We forked and modified Paperclip to ship a hosted multi-tenant product with pre-configured operator templates, approval queues, billing, and onboarding. The underlying framework remains open source; Pocket Square's contribution is the practitioner curation, hosted deployment, and the founder-friendly chrome on top.
- Why a Chief of Staff and not just one big AI?
- Single point of contact, fluent in your business, who briefs the specialist team, reviews their output, flags what needs your eyes, and comes back with answers. You manage one role, not five tools.
- Do my AI operators have access to my data?
- Each tenant runs in an isolated Paperclip instance with its own database. No cross-tenant data leak at the container or DB level. Approvals gate any high-risk action before it executes, and execution logs surface in a plain-English timeline you can audit.
- Can I use my own model API keys?
- BYOK is on the roadmap. Today the credits-based pricing covers OpenRouter access to the major model families (Claude, GPT, Gemini) so you can swap models per operator without managing multiple subscriptions.
- When can I start?
- Join the waitlist for early access. We're onboarding founders in cohorts so we can stay close to feedback during the early months.
Still have questions?
Join the waitlist →