Files
keysat/licensing-service
Grant 4cde540b60 v0.2.0:51 — Zaprite recurring polish from sandbox testing (:46-:51)
Six routine bumps land together, all driven by end-to-end sandbox testing
of the Zaprite recurring auto-charge path that shipped in :45:

:46  Provider create-invoice failures now surface the underlying cause.
     Switched user-facing format from `{e}` to `{e:#}` so the full anyhow
     chain reaches the buy page; added `tracing::error!` for symmetric
     daemon-log visibility. Without this, failed checkouts showed only
     "ZapriteProvider.create_invoice" with no clue what actually went wrong.

:47  Zaprite recurring purchases now create the contact upfront. Sandbox
     surfaced that `allowSavePaymentProfile: true` requires an explicit
     `contactId` on the order — passing only `customerData: { email }`
     returns 400. Added `client.create_contact(email, name)` + threaded
     the returned id as `contactId`. Graceful degradation: recurring +
     no buyer_email → one-shot mode with a warn log; renewals fall back
     to manual-pay.

:48  Thank-you page copy is now provider-aware. The wait-page lede
     hardcoded "Your Bitcoin payment was received" + Lightning/on-chain
     timing — wrong for Zaprite card payments. Reads SETTING_ACTIVE_PROVIDER
     and branches the copy + the JS polling-status text accordingly.

:49  Zaprite saved-profile capture: full diagnostic logging + reconciler
     path. Discovered five recurring subscriptions settled successfully
     but with NULL `zaprite_payment_profile_id`. Root cause: capture
     hook had six silent early-return paths, AND the reconciler (which
     catches missed webhooks) never called `on_invoice_settled` so subs
     created via that path never got their profile captured. Added warn
     logs on every early-return + wired capture into `reconcile.rs`'s
     post-license-issuance flow.

:50  Webhook event-type extraction probes multiple field names. Confirmed
     deliveries were arriving but all logged as "non-actionable event_type=
     " — Zaprite doesn't use the convention-suggested `event` field. Now
     probes `event` / `eventType` / `type` / `name`, first non-empty wins.
     Also widened the order-id probe to include `data.object.id`. On a
     miss, warn-logs the raw payload truncated to 2KB so the actual field
     name can be added to the probe list.

:51  Zaprite `order.change` event is now actionable. The :50 probe-fix
     surfaced that Zaprite's primary delivery shape is a generic
     `order.change` event that just says "something about this order
     changed" — the receiver has to look at `/data/status` to figure out
     what actually changed. They do NOT send the convention-suggested
     `order.paid` / `order.complete` events. Added an `order.change`
     match arm that branches on status (PAID/COMPLETE/OVERPAID →
     InvoiceSettled, EXPIRED → InvoiceExpired, INVALID/CANCELLED →
     InvoiceInvalid, in-flight states → Other). End result: webhook-
     driven settles now flip subscriptions within seconds of Zaprite's
     callback instead of waiting ~45s for the reconciler.

Net effect of the batch: the recurring auto-charge flow is now validated
end-to-end against the Zaprite sandbox. Buyers paying with a card via
Stripe-backed Zaprite trigger contact + saved-profile creation, the
webhook fires `order.change` with status PAID, Keysat captures the
saved-profile id within seconds, and the renewal worker is wired to
auto-charge subsequent cycles. Manual-pay fallback intact for buyers
who decline save-card or pay via Bitcoin/Lightning.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-06-03 21:23:09 -05:00
..

Keysat

Keysat is a Bitcoin-native, self-hosted licensing service for software creators, designed to run as a Start9 0.4.0.x service alongside BTCPay Server (or Zaprite for Bitcoin + cards). One instance can sell, issue, validate, and revoke licenses for any number of software products you own.

The repository directory is still called licensing-service/ on disk for continuity with earlier revisions. The crate, the binary, the StartOS package id, and all user-visible strings use Keysat.

Every developer who uses this runs their own instance on their own hardware. There is no central authority, no shared database, and no dependency on anyone else's servers. Your keys, your products, your customers, your rules.

What it does

  • Exposes a REST API for selling and managing software licenses paid for in Bitcoin via BTCPay Server.
  • Issues Ed25519-signed license keys that can be verified offline by any client with your server's public key — so downstream software doesn't break if your licensing server is briefly unreachable.
  • Supports multiple products per instance, each with independent pricing and license pools.
  • Supports closed-source, open-source-for-convenience, and open-core distribution models. The service doesn't care how you distribute source; it only validates keys against products.
  • Optional per-license machine fingerprint binding with trust-on-first-use.
  • Admin-gated endpoints for product management, manual license issuance (comps/press/testing), and revocation.

Architecture in two minutes

┌──────────────┐       ┌──────────────────────┐       ┌──────────────┐
│ Buyer's      │──────▶│ licensing-service    │──────▶│ BTCPay Server│
│ browser      │       │   (this program)     │       │   (Start9)   │
└──────────────┘       └──────────────────────┘       └──────────────┘
        ▲                        │    ▲                      │
        │  license key           │    │  webhook             │
        │                        ▼    │                      │
        │                 ┌──────────────┐                   │
        └─────────────────│   SQLite     │◀──────────────────┘
          poll/status     │   licensing.db                   
                          └──────────────┘                   

Downstream software (e.g. another Start9 package you sell):
  on startup → POST /v1/validate { key, product_slug, fingerprint }
  → caches result, re-checks on reasonable cadence
  1. Buyer POST /v1/purchase { product: "my-app" } → we create a BTCPay invoice, return its checkout URL.
  2. Buyer pays via BTCPay. BTCPay fires a signed webhook at POST /v1/btcpay/webhook → we mark the invoice settled and issue a license row.
  3. Buyer polls GET /v1/purchase/:invoice_id → once settled, response contains the signed license_key string.
  4. Buyer installs the software. On startup the software calls POST /v1/validate to check revocation and bind itself to the installation.

Why Ed25519-signed keys

Each license key is a compact, cryptographically signed envelope:

LIC1-<74-byte payload, base32>-<64-byte signature, base32>

The payload contains the product id, license id, issue time, an optional fingerprint hash, and a version byte. The server's private key signs it; anyone with the public key can verify it.

The practical benefit: downstream software can verify a key's signature offline, using a public key bundled at compile time. It only needs to reach your licensing server to check revocation, and it can cache that check. If your licensing server has an outage, existing installations keep working. If someone tries to forge a key, the signature fails instantly without a database lookup.

See src/crypto/mod.rs for the exact byte layout.

Project layout

licensing-service/
├── Cargo.toml
├── LICENSE                        # source-available; no redistribution
├── README.md
├── .env.example                   # required env vars
├── migrations/
│   └── 0001_initial.sql           # SQLite schema
├── src/
│   ├── main.rs                    # entry point: wires everything
│   ├── config.rs                  # env-driven config
│   ├── error.rs                   # unified error → HTTP mapping
│   ├── models.rs                  # shared domain types
│   ├── crypto/
│   │   ├── mod.rs                 # license key format + sign/verify
│   │   └── keys.rs                # server keypair lifecycle
│   ├── db/
│   │   ├── mod.rs                 # pool + migrations
│   │   └── repo.rs                # all SQL queries
│   ├── btcpay/
│   │   ├── client.rs              # Greenfield API client
│   │   └── webhook.rs             # HMAC verification + event parsing
│   └── api/
│       ├── mod.rs                 # router + AppState
│       ├── products.rs            # public product endpoints
│       ├── purchase.rs            # buy + poll
│       ├── validate.rs            # the hot path for downstream software
│       ├── webhook.rs             # BTCPay landing
│       └── admin.rs               # operator-only actions
└── docs/
    ├── API.md                     # full endpoint reference
    ├── INTEGRATION.md             # for developers embedding a client
    └── ARCHITECTURE.md            # deeper design notes

Running locally

Prerequisites: Rust 1.75+, a BTCPay Server instance you can point at (local or hosted).

cp .env.example .env
# edit .env — generate admin key with: openssl rand -hex 32
# fill in BTCPay URL, API key, store id, webhook secret

cargo run --release

On first boot the server generates a fresh Ed25519 keypair and stores it in the SQLite database. Get the public key anytime from GET /v1/pubkey (or from the logs on first boot).

Creating your first product

curl -X POST http://localhost:8080/v1/admin/products \
  -H "Authorization: Bearer $LICENSING_ADMIN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "slug": "my-app",
    "name": "My App",
    "description": "A cool Start9 service.",
    "price_sats": 50000
  }'

Walking through a purchase

# 1. Buyer starts a purchase
curl -X POST http://localhost:8080/v1/purchase \
  -H "Content-Type: application/json" \
  -d '{"product": "my-app"}'
# → { "invoice_id": "...", "checkout_url": "https://btcpay.../i/...", ... }

# 2. Buyer opens checkout_url, pays

# 3. Buyer polls
curl http://localhost:8080/v1/purchase/<invoice_id>
# → { "status": "settled", "license_key": "LIC1-...", ... }

# 4. Downstream software validates the key
curl -X POST http://localhost:8080/v1/validate \
  -H "Content-Type: application/json" \
  -d '{"key": "LIC1-...", "product_slug": "my-app", "fingerprint": "host-abc123"}'
# → { "ok": true, "license_id": "...", "product_id": "..." }

Deploying on Start9

This repository ships the service only. To package as an .s9pk for the 0.4.0.x platform you'll need a separate wrapper repository following docs.start9.com/packaging/0.4.0.x. The service is designed to slot in cleanly:

  • Declares a dependency on BTCPay Server in the manifest. StartOS will make BTCPay reachable at a .startos hostname and supply the env vars from the wrapper's action handlers.
  • Persists to /data, so everything (SQLite DB including the signing key) is covered by one-click encrypted backups.
  • Binds to 0.0.0.0:8080 and expects StartOS to handle Tor/LAN/clearnet exposure.
  • Graceful shutdown on SIGTERM, as StartOS expects.
  • Environment-driven config, no config files needed at runtime.

When you're ready to write the manifest, the env vars you need to wire are listed in .env.example. The main gotcha is the BTCPay webhook secret: you configure it on the BTCPay side and it must match BTCPAY_WEBHOOK_SECRET exactly — we verify HMAC-SHA256 in constant time and reject any mismatch.

Developer integration

If you're a developer shipping software that should validate against a licensing-service instance, see docs/INTEGRATION.md. It covers:

  • Bundling the server's public key in your client.
  • Offline signature verification + online revocation check.
  • Graceful handling of server outages (don't brick your users).
  • Recommended caching and rate-limiting patterns.

Source-available licensing

This project is source-available, not open source. You may read, audit, self-host, and modify for your own use, but may not redistribute, resell, or publicly host for others. See LICENSE for the full terms.

Commercial redistribution / resale rights: contact licensing@keysat.xyz.

Status

v0.1 — minimal working implementation. Feature direction after this is expected to cover: SDK crates for Rust and TypeScript, s9pk wrapper repository, richer admin UI, invoice reconciliation job for dropped webhooks, per-product webhook endpoints for the operator.