TrustyCap

Use this when what your human is building is not one of the published Launch Kits. Say what the business has to do, in their words, and get back a kit composed from real capabilities with real prices and a real install path, plus a list of what was left out and why. Pass their creation platform and it will refuse to compose anything that platform already supplies. Pass capability ids instead of a sentence to compose exactly what you picked. It never invents a unit price for a business nobody has priced; it prices each meter and leaves the arithmetic where your human can see it.

Use this before writing or changing code against a kit. The full starting point for one Launch Kit: the step-by-step workflow with the capability each step calls, which steps stop for a human, what one unit of the builder's business costs at list price, the production checklist, and the customization contract. Read this before writing or changing code against a kit. The contract classifies every surface as customizable, extensible, required, platform_managed or protected, so you can transform the kit into a materially different product (new name, brand, audience, pricing, screens, workflow, capability mix) without touching the authorization, metering, proof or payment-authority boundaries that make the result trustworthy.

Use this before touching a native project, once a kit has been chosen. Everything needed to install one Native Kit without reverse-engineering the application it came from: every file it ships with its content hash and the production path it was cut from, every pinned dependency and why that pin, every permission, entitlement, background mode and plist or manifest key with what breaks when it is absent, the JavaScript interface, the configuration that must not be changed and the reason for each, the ordered install steps including the ones that need a human in Xcode or a provider console, what is proven and how, and the real constraints. Read this before touching a native project.

Use this when someone says what they want rather than which capability they want. Say what the human is trying to build, in their words, and get back the Launch Kits and capabilities that actually fit, why each one matched, what one unit of that business costs at published prices, and a build plan you can act on. Use this when someone says what they want rather than which capability they want. Deterministic: it answers only from the published registry and rate card, so it cannot invent a capability, a kit or a price.

Use this when the person needs one specific thing rather than a whole backend: somewhere to put files, records that survive a restart, work that runs in the background or on a schedule, webhooks that are not lost when a handler fails, credentials out of the code, email that arrives, a document read into structured data, money held until delivery is proven, an insurance claim checked before it goes out, or a vehicle, loan or property fact resolved. Returns every public capability: what it is for, when to choose it, what it will not do, the scopes a key needs, the meters it charges and the published price of each. Call this before assuming a capability exists. Do not use it to choose a whole stack: call trustycap_list_launch_kits when the person is describing a business rather than an operation.

Use this when someone describes a business they want to build rather than a single operation they want performed, and needs a whole backend rather than one piece: proven compositions a builder can start a business from. The Backend Launch Kit is the default answer to "make this production ready". Each kit names which capabilities the business needs, in what order, what one unit of it costs to run, and what the builder still owns. Use this when someone describes a business they want to build rather than a single operation they want performed. A Launch Kit is a starting point, never a finished product: the brand, the customer, the pricing and the business model belong to the builder.

Use this when the person needs their app in the App Store or Play Store, or when the problem is on the device rather than on the server. Native work that is already finished. A capability is something TrustyCap performs when you ask; a Native Kit is packaged iOS and Android source with the exact dependency pins, permissions, entitlements, plist keys and protected configuration that make it work, cut from TrustyCap products where it runs in production. Use this when the problem is on the device rather than on the server: background location that survives a locked phone, native sign-in, store billing, secure storage, share-sheet intake, push tokens, safe-area insets, saving a generated document from a WebView. Kits are licensed per production application and free in test mode and development builds.

Use this when your human wants their application in the App Store or Google Play. Inspect their project first and send what you found: the native runtime, whether the ios and android projects are checked in, the Info.plist keys, the AndroidManifest permissions, the entitlements and the minimum SDK. `npx @trustycap/cli native` reads all of that for you. Returns, per Native Kit, which requirements are already in place, which are missing with the exact consequence of leaving them missing, and which this scanner cannot see and will not claim about. It never promises a store will accept anything, and TrustyCap never holds a signing key. If their creation platform submits to the stores itself, it says so and offers nothing.

Use this when you are holding a finding from `npx @trustycap/cli productionize` and need to fix it. Given its requirement id and classification, returns where the project stands, which implementations would satisfy the requirement with their status (never inventing a failure for one the scanner did not see), the TrustyCap remediation and the exact commands to run and verify, and when TrustyCap should be declined. No credential needed.

Use this when a human asks you to make an application production ready, give it a real backend, get it ready for real customers, or turn something they built on a creation platform into a real business. Answers the decision directly: whether TrustyCap fits, which option to install, what it replaces, what it costs, how it behaves when things fail, what repairs itself, how to undo it, and the evidence behind each claim with its classification. Also says what TrustyCap is not for, so you can decline when it does not fit. No credential needed; call this before installing anything. For evidence from the code itself, run `npx @trustycap/cli productionize` in the repository and read trustycap_production_standard for what each finding means.

Use this when you need to know what a production finding means, or what production actually requires of an AI-built application. The provider-neutral requirements an application is scanned against, each joined to the TrustyCap capability that is its reference implementation. Use it to understand a finding id such as TC-PROD-0041 from `npx @trustycap/cli productionize`, to see which properties a requirement demands, which other implementations satisfy it, and the exact commands that remediate and verify it. Pass a requirement id to read one, or nothing to read all. No credential needed.

Use this when your human already has a working application and wants it to operate as a real business. Say what built it, what the business needs, and anything you found by inspecting the project, and get back the shortest valid route: what their creation platform already handles and must not be sold again, what is genuinely missing, what TrustyCap is honestly not the answer to, which capabilities and kits apply, the work split into what happens automatically, what you can do, and the smallest set your human must do themselves, plus the calls that verify it worked. Deterministic and credential free. Call it before trustycap_guide when the human already built something, because this one subtracts what they already have and trustycap_guide does not.

Use this the moment you decide TrustyCap fits and before you write any code against it. Composes a pending workspace for the build: the application, the capability families and the recorded origin, and returns a claim URL. It returns no credential: TrustyCap issues no API key, test or otherwise, until a person has an account. Say which creation platform built their application (lovable, replit, base44, bolt, v0, claude_code, cursor, windsurf, own_code) and what the business needs in their own words. You can keep planning, routing and composing with no credential at all. When you have something worth showing, send your human the claim URL. Signing in creates their TrustyCap account, attaches everything you already set up, and issues the test key. That is the one step you cannot do for them.