# Sightspool

> Go from what to why.
>
> You know where users drop. Not why. Sightspool asks the people who dropped —
> one real answer from a real user beats a month of dashboards — and records what
> you expected, so you never investigate the same question twice.
> Sightspool connects approved product, customer, commercial and repo context
> to one governed product-intelligence agent, with senior UX judgement available
> when the evidence or stakes require it. Sightspool makes product use measurable,
> surfaces falsifiable assumptions (assumption -> question -> signal -> threshold
> -> verdict), applies a visible proof gate, proposes client-owned action and
> returns to learn what happened. Core invariants: no proof, no post (every metric
> cites its live source); the AI proposes, the client decides; an agent never
> rules an assumption "verified"; below minimum sample the register says "still
> gathering", never a verdict.

## Docs

- [From audit to answers](https://www.sightspool.com/docs/from-audit-to-answers): You ran the free ux-assumptions audit and got a file. Here is what it is, what to do with it next, and what changes if you bring it into Sightspool.
- [Getting started](https://www.sightspool.com/docs/getting-started): Explore the demo workspace, create your own, run a first product read and own the action through outcome review.
- [Connecting sources](https://www.sightspool.com/docs/connect-sources): What Sightspool reads, how your credentials are stored, and why sample data can explore the product but never back a published finding.
- [The assumption register](https://www.sightspool.com/docs/the-register): How an assumption becomes falsifiable — question, signal, threshold, verdict — and how ruled verdicts write back into your coding agent's context file.
- [Staying current](https://www.sightspool.com/docs/staying-current): How the register keeps up with a moving product: repo re-arming, the weekly digest, your recorded dispositions, and verdict write-back into your coding agent.
- [Demand probes](https://www.sightspool.com/docs/demand-probes): Test demand for a feature that doesn't exist yet — a dashed-border button where the feature would live, honest by construction, counted as votes in your register.
- [Coding-agent access (MCP)](https://www.sightspool.com/docs/coding-agent-access): Connect Claude Code, Cursor or any MCP client to your workspace so your coding agent can check what's known — and deposit the assumptions it makes while building.
- [Deep-capture studies](https://www.sightspool.com/docs/deep-capture): Bounded, cohort-scoped session recording in your own PostHog — justified by one assumption, approved by you, armed for a set window, then switched off again.
- [Expert reads and human escalation](https://www.sightspool.com/docs/expert-reads): How Sightspool answers a question when there is not enough traffic to measure anything — and what actually happens when work escalates to a senior human.
- [Research studies](https://www.sightspool.com/docs/research-studies): When behaviour cannot explain itself: research briefs, in-product interviews, PostHog-delivered surveys, and how a finished study becomes citable evidence.
- [Honesty guarantees](https://www.sightspool.com/docs/honesty-guarantees): The claims Sightspool structurally cannot make: unsourced numbers, verdicts below sample, agent-declared certainty, actions you didn't approve, or an undisclosed painted door.

## SDK

- [SDK reference](https://sdk.sightspool.com): @sightspool/sdk — in-product
  capture (intent + effort + account), served interventions, and privacy
  controls. API reference, versioned with the package.

## Product

- [Product](https://www.sightspool.com/product): the operating layers and boundaries.
- [Senior UX review](https://www.sightspool.com/ux-studio): the focused human review included after a founding user completes an independent first read.
- [How it works](https://www.sightspool.com/how-it-works): the question-to-outcome loop.
- [About](https://www.sightspool.com/about): the company thesis and the founder's assurance role.
- [Founding beta](https://www.sightspool.com/founding-beta): join the open 30-day learning beta,
  bring one real product question and receive one focused senior UX review
  after your first independent run.

## Journal

- [The person at the other end of the build](https://www.sightspool.com/blog/the-person-at-the-other-end-of-the-build): Software can now be built at extraordinary speed. Sightspool exists to keep its assumptions, evidence and human consequences visible after it ships.