Project Detail

Araldo

A developer-first social distribution API that renders, validates, schedules, and publishes one announcement across thirteen platforms — with a built-in MCP server so AI agents can publish safely on a brand’s behalf.

Positioning

A complete multi-tenant SaaS — API, dashboard, CLI, MCP server, billing — designed for both developers and AI agents as first-class callers.

Category Developer API and SaaS

Status Early access

Website https://araldo.dev

Overview

What this project demonstrates

Araldo takes one announcement from a product and handles the rest: it renders the post for each platform from a schema-validated template, checks it against that platform’s rules, schedules it into the brand’s weekly slots, publishes it, and reports the outcome by signed webhook. It never posts twice — when a delivery is ambiguous it retries only where retrying is safe and otherwise asks a person. Around that core sit newsletters, engagement and signup analytics, ad-spend reporting, and monthly brand reports. The server is open source under the AGPL; the hosted product adds sign-up and Stripe billing. Spectrum Labs’ own products publish through it.

Core framing The publishing layer products and AI agents can trust

Good Fit For

Products that need to integrate with many third-party APIs reliably

Developer platforms that need a clean API, SDK contract, CLI, and webhooks

AI agent integrations via the Model Context Protocol

Multi-tenant SaaS needing organizations, roles, SSO, and billing

Core Delivery Areas

Third-party API adapters with per-platform validation, media handling, and rate-limit holds

Exactly-once publishing semantics with safe retries and a human-review path

API platform conventions: idempotency keys, cursor pagination, problem-details errors, signed webhooks

Multi-tenant security: scoped API keys, passkeys, TOTP, OIDC SSO, and per-org envelope encryption

Platform Stack

Single Go binary with Postgres as the only required dependency

Server-rendered dashboard with Tailwind that works without JavaScript and meets WCAG 2.2 AA

OpenAPI contract kept in step with the routes by tests, published under Apache-2.0 for client generation

Kubernetes via Helm, Cloudflare Tunnel and Access, and OpenBao Transit for the master key

AI Pipeline

MCP server over stdio and HTTP, scoped to what the caller’s API key allows

Preview returns every rule violation with a stable code, so an agent can fix its own draft and retry

Sandbox channels and test-mode keys let agents experiment without reaching real audiences

Applied Product Capabilities

Thirteen platforms, including Bluesky, Mastodon, X, LinkedIn, Threads, Instagram, YouTube, TikTok, Discord, and Telegram

Slot-based scheduling with approvals, rescheduling, and deadlines

Newsletters handed off to the email provider with approval and test sends

Engagement, signup attribution, and ad-spend analytics with monthly brand reports

CLI with terminal sign-in and local webhook streaming

Hosted plans with Stripe Checkout, customer portal, and per-org limits

What A Client Could Hire Spectrum Labs To Do

Build a developer-facing API product with a contract, CLI, webhooks, and docs

Make an existing product usable by AI agents through an MCP server

Integrate many third-party platforms behind one reliable interface

Take an open-source core to a hosted, billed SaaS

What’s Next

Automated daily verification of every adapter against real accounts

Paid promotions with per-brand spend caps, and AI drafting with the organization’s own model key