Spiritt

Before anything else: the Spiritt listed in most tool directories — the AI business-planning and founder-networking platform with guided templates, market analysis dashboards, financial modeling, and a mentor community — is not what you get when you visit spiritt.ai today. The live product has moved. Spiritt now presents itself as an AI startup studio and autonomous agent workspace: describe a product and its agents architect, build, deploy, and then maintain it, on a subscription. If you arrived here looking for business-plan templates and networking, that offering has been folded into, or replaced by, a done-with-you build service. I am reviewing what the site actually sells now, and rating the trust signals accordingly, because a listing that describes a product the vendor no longer shows is a listing that will waste your afternoon.

What it actually does

There are two faces to Spiritt now, and they share one funnel.

The first is the startup studio, laid out on spiritt.ai/apps. The entry point is a free Blueprint plan: Spiritt scopes a first version with you, defines the MVP, the screens, the features, entity and user flows, and a brand identity, and hands back a professional PDF app plan plus a consultation call with a product specialist — at $0, explicitly 'completely free.' It is a lead-generation artifact, and a sensible one: it commits the vendor to a scoping conversation before anyone pays.

The paid tiers are service-as-software. Managed, at $350/month at time of writing, promises a fully functional custom product (backend and frontend), a visual Canva-like editor for later changes, $400 of premium AI and compute credits, mobile and desktop apps, a payments system with split payments and subscriptions, hosting on scalable servers, a personalized brand book, and end-to-end QA — with the pitch that a production-grade, cross-platform app ships 'in under a month' with no setup fee. The Startup Program, at $500/month and acceptance-gated, adds priority support with a stated SLA, a dedicated account executive, $600 of AI and compute credits, enterprise-grade QA, business and marketing help, early beta access, a full launch plan, and unlimited page count. There is also an accelerator partner program that hands cohort founders credits and a hands-on workshop.

The second face is the agent platform on spiritt.io, marketed as an autonomous workforce: an agent that plans, writes software, deploys to production, connects to real services, runs background jobs, monitors what it shipped, and fixes its own breakage. The site's demo transcript shows the agent shipping a multi-tenant B2B sales-intelligence platform — database schema, backend, LLM scoring agents, five data integrations, outreach sequencing, Stripe billing, load tests, a three-region deployment — in roughly 1.5 hours of agent time. That transcript is a vendor demonstration, and I read it as one: it shows what the vendor wants the ceiling to look like, not what your first afternoon will look like.

The vendor's FAQ positions it against adjacent categories: against chatbots (it is an active workspace, not a reactive one), against Claude Code (cloud workspaces and autonomous operation versus local CLI development), and against Lovable/Replit/Base44 (full-stack builds with maintenance and self-optimization versus front-end prototyping). The comparisons are the vendor's own, but they do tell you where Spiritt wants to sit: above app builders, beside coding agents, as an operating layer for a company.

My testing notes

Two observations from walking the public surface and cross-checking it against directory records.

First, the pricing as published is service pricing wearing software pricing's clothes, and the free Blueprint is the only piece I could verify at zero risk. Managed at $350/month for a custom backend-and-frontend product with payments, hosting, QA, and mobile/desktop apps is either extraordinary value or a subscription-shaped retainer whose scope tightens once you are in — and the public material cannot distinguish the two. Concrete things I checked: the plan bullets on spiritt.ai/apps are specific (split payments, brand book, E2E QA, unlimited page number gated to Startup Program), the funnel goes through an application at joystick.spiritt.io rather than instant checkout, and the Startup Program explicitly 'requires acceptance into the program.' Application-gated pricing plus generous promises is a pattern I have seen from agencies before; it is not a red flag by itself, but it means the honest evaluation unit is the free Blueprint deliverable and the consultation call, not the marketing page. I would treat the Blueprint as the trial: if the scoping PDF is sharp and specific, that is evidence about the machine that produced it.

Second, the trust signals around Spiritt are thin and partly self-referential. One directory snapshot (try.tools, June 2026) recorded minimal GitHub activity — three stars — and a modest site authority score, and explicitly framed its own summaries as AI-generated impressions. The vendor claims its agents are 'award-winning' and 'self-improving,' and that it runs 'the vast majority' of its own software on its own technology ('SPIRITT is built on SPIRITT') — all vendor claims with no third-party verification I could find. There are no independent hands-on reviews of the current agent platform in my searches; the directory coverage that exists describes the older planning-and-networking product and clearly predates the pivot. The old product's numbers survive in those listings — a premium tier at $29.99/month, templates, mentorship — and none of them appear on the live site. When the evidence base for a product is its own landing pages plus AI-generated directory copy, the correct posture is provisionally skeptical.

To its credit, the operational transparency is better than most agent-workspace pitches: the FAQ is candid that finance output is 'a powerful assistant for finance operations, not a licensed financial advisor,' that changes can be rolled back when the agent errs, and that you can read its logs and edit the code yourself. Those are the answers of someone who has thought about the failure modes.

What it does not do — and what I cannot verify

I could not verify, from any independent source: the 'in under a month' build timeline, the 1.5-hour full-platform build in the demo transcript, the claim of self-healing production infrastructure, or the 'award-winning' agent credentials. The old planning-and-networking toolkit — templates, market analysis dashboards, financial modeling, mentor matching — is absent from the current site; if that is your need, look at the directory-era descriptions only as history. There is no public API (try.tools records API availability as 'No'), no self-serve tier for the agent platform with published limits, no published model/credit economics for the workspace itself, and no public status page, uptime history, or third-party security attestation that I could locate. The $400 and $600 'premium AI and compute credits' bundled into the paid tiers have no published exchange rate against real usage, so their face value is unverifiable. And because entry runs through an application rather than a checkout, price discipline depends on the contract you sign, not the page you read.

Verdict

Spiritt is best understood as a small, ambitious studio selling scoped product builds and an agent workspace under one subscription, not as the startup-planning toolkit the directory listings describe. The free Blueprint is a low-risk way to test the substance behind the pitch, and the FAQ shows unusual candor about limits. But the claims are large, the independent evidence is close to zero, the pricing's value hinges on scope I cannot see, and the product's identity changed recently enough that most of what is written about it is stale. I would engage for the free scoping pass and treat any paid commitment as a services engagement requiring a written scope. Rated 2.7: unverified functional depth behind broad claims and a departed planning toolkit (2.6), paper value that is attractive but contract-dependent (3.2), an application funnel and hands-on consultation that are approachable though not self-serve (2.8), accelerator partnerships and named support roles as the ecosystem play (2.8), and weak reliability signals — pivot-era confusion, near-zero community footprint, no independent reviews (2.0).

Sources

• https://spiritt.ai/
• https://spiritt.ai/apps
• https://spiritt.ai/accelerators
• https://spiritt.io
• https://joystick.spiritt.io/apply-to-spiritt
• https://store.hypergpt.ai/spiritt
• https://findaitools.app/tools/spiritt-ai
• https://fastd.top/ai-tools/spiritt
• https://try.tools/tools/spiritt
• https://qoo.im/tools/spiritt