The public API documentation for Day3, served at docs.day3.app.
Source-available, not open source. This code is published for transparency and reference. See LICENSE for terms.
Day3 is marketing and transactional email for small SaaS teams, billed by sends rather than by list size. These docs are the reference for talking to it over HTTPS.
- Quickstart, authentication, conventions, errors: the things that apply to every endpoint
- One page per resource: emails, contacts, audiences, campaigns, suppressions, automations
- Integrations: signed webhooks, and the MCP server that lets Claude Code or Cursor draft a campaign
- Guides: task-shaped walkthroughs, including migrating a list off another platform
Nextra 4 on Next.js 16, statically exported and hosted on GitHub Pages. Same stack as exit1.dev.docs.
| Website | day3.app: what Day3 is, pricing, and the case for it | Repo |
| App | go.day3.app: sign in, write campaigns, send email | Repo |
| Documentation | docs.day3.app: API reference, guides, webhooks, MCP | Repo |
Built by Pradsgaard Labs. Also from the same workshop: exit1.dev, uptime monitoring.
npm install
npm run dev # http://localhost:3000npm run build # next build -> out/, then Pagefind -> out/_pagefindnext build exports the whole site to out/, then Pagefind indexes the
exported HTML in place. Indexing the export rather than .next/server/app
means the search index always matches what actually shipped, and nothing
binary needs committing.
Push to main. .github/workflows/deploy.yml
builds and publishes out/ to GitHub Pages.
The custom domain comes from public/CNAME. public/.nojekyll
is insurance: both _next/ and _pagefind/ start with an underscore, which a
stray Jekyll pass would drop.
One page per resource under app/, as page.mdx. Sidebar order and
titles come from the _meta.js file in each directory.
Pages are grouped by resource rather than one page per endpoint. That is deliberate: roughly 40 operations split one-per-page would be 40 thin pages, and thin pages on a new domain tend to sit in Google's "Discovered, currently not indexed" bucket instead of ranking.
| Path | Covers |
|---|---|
app/page.mdx |
Introduction |
app/quickstart, app/authentication, app/conventions, app/errors |
The things that apply everywhere |
app/emails … app/suppressions |
One page per resource, including app/automations |
app/webhooks, app/mcp |
Integrations |
app/guides |
Task-shaped walkthroughs |
Everything here is derived from the app repo
(Mopra/day3.app, checked out locally as
../day3.app), which is the source of truth:
| Source | Gives us |
|---|---|
src/lib/api-docs.ts |
buildReferenceMarkdown(), the canonical reference. test/api-docs.test.ts asserts every app/api/v1/**/route.ts appears in it, so it cannot fall behind the code. |
src/api/v1/serialize.ts |
Exact response shapes. Every field in these docs is a real serializer field. |
src/api/v1/errors.ts, scopes.ts, auth.ts |
Error codes, the two scopes, key format. |
docs/api-v1-spec.md |
Design rationale. Parts of §1 are stale: it predates scopes. |
docs/webhooks.md |
Receiver contract and the signature algorithm. |
src/mcp/tools.ts |
The MCP tool list. |
src/lib/plans-catalog.ts |
Plan limits: the free tier's 500-subscriber cap and the 100-email sandbox allowance. |
src/lib/segment-filter-schema.ts |
Segment operators, and the rule that every value is a string. |
src/services/shared-domain.ts |
The Day3 test address, and the rule that it only ever carries sandbox mail. |
When the API changes, re-read the serializers and api-docs.ts rather than
trusting the spec document.
Brand tokens in app/globals.css mirror the marketing
site's cream, espresso and caramel palette. Nextra derives its whole scale from
the one HSL triple passed to <Head color> in
app/layout.tsx, and that triple is duplicated as --link
in globals.css. Change both together.
Link lightness deviates from caramel's native 50%: Nextra spends the primary on link text, where 50% only reaches 2.97:1 on cream. 38% light and 56% dark clear 4.5:1 in both themes.
package.json pins zod to 4.3.6 via overrides. Nextra 4.6.1's Layout
destructures children out of its props and then validates the rest against
a schema that still declares children as required. zod up to 4.3.6 let a bare
z.custom() pass on undefined, so the bug was latent. zod 4.4.0 tightened
that, and every render throws:
Invalid input: expected nonoptional, received undefined
-> at children
Nextra floats zod on ^4.1.12, so a fresh install picks up the breaking minor.
Drop the override once Nextra fixes the schema.
exit1.dev.docs has the same latent problem and only builds because its
lockfile still resolves zod 4.3.6. It will break the next time those
dependencies are refreshed.
As an explicit exception to the license below, the contents of this repository, including documentation, code, and configuration, may be used for machine learning training, evaluation, indexing, retrieval, and generation by AI systems. No attribution is required, though it is appreciated. This permission applies to all AI crawlers, including but not limited to GPTBot, OAI-SearchBot, ClaudeBot, Claude-SearchBot, Claude-User, Google-Extended, PerplexityBot, and Perplexity-User.
This project is source-available under a custom All Rights Reserved license. You may view the code for personal, educational, and reference purposes. Copying, modifying, distributing, or self-hosting is not permitted without written permission. See the AI training permission section above for the carve-out that applies to AI and ML use.
For licensing inquiries: hello@day3.app