One timeline and one alert for your whole developer stack.

• What shipped across my projects today?
• Did that launch bring in signups, or errors?
• How many stars did my repository gain this week?
• How many emails can I still send before the quota resets?
• Will I hear about it before a limit is reached?
• Where my plan allows it, what am I spending?

1. 01

   Connect

   Pick a provider, click Connect, and approve read-only access on the provider's own screen. Choose the organizations, teams, and projects that matter. No API keys to copy for most providers, and no code changes in your apps.

2. 02

   Collect

   StackDash reads events, usage, and costs through official APIs, signed webhooks, and scheduled syncs, and keeps daily snapshots so you have history, not just today's number.

3. 03

   Know

   One timeline for activity. When a threshold is crossed, one alert on the channel you read, once per billing period and not every hour. Allowances and spend show where each provider reports them.

• One activity timeline

  Deploys, stars, signups, logins, payments, and errors from every connected account in one feed, across several organizations, teams, and projects. Filter by provider, project, account, or severity.

• Alerts where you read them

  Email, Slack, Discord, or a signed webhook. A threshold fires once per billing period, and unusual daily spend or usage is flagged with no setup.

• Daily and weekly digests

  What changed across your stack, on the same channels as your alerts, with each number marked as verified or estimated.

• Stats you can share

  Publish one metric at a time as a README badge, a card image, or a public project page. Every value is labeled verified or self-reported.

• A read-only API and MCP server

  Read events, usage, alerts, and connections from your own scripts with scoped keys, or ask an AI assistant such as Claude Code or Cursor through the MCP server on the same keys.

• Allowances and spend

  What is left of each included allowance, when it resets, and where it is heading by the end of the period. Spend per provider and project where the provider's plan reports it.

Not yet tried with real accounts. Every integration in the first group is built from the provider's documentation and tested against sample responses taken from it. None has been connected to a real account yet, so each provider's page in the docs lists what is still unverified. Until Sentry approves the hosted integration for publication, only the organization that registered it can install it.

About ChatGPT and Claude subscriptions. Plans like ChatGPT Plus or Claude Pro are separate from API billing, and most do not offer a documented usage API. StackDash will only show subscription usage where an official route exists. It will never scrape your sessions or ask for your password to fill the gap.

• Read-only wherever the provider allows it

  OAuth first, with the narrowest read-only scopes the provider offers, so most connections need no API key at all. Where a provider has no OAuth route, a restricted read-only key is the fallback. Write permissions are never requested for a dashboard that only reads.

• Permissions shown before you connect

  Every integration is labeled Read-only, Restricted API key, or Elevated access required. If a provider only shares usage data with broader access, as Resend's OAuth does, StackDash tells you before you connect and never calls it read-only.

• Credentials stay on the server

  Tokens are encrypted at rest, never sent to the browser, never written to logs, and isolated per workspace. Disconnect any service whenever you like.

• No scraping, no passwords

  Only official APIs, authorized integrations, and signed webhooks. Where no official route exists, StackDash says so instead of guessing.

• Self-host it

  Run the whole thing yourself with Docker Compose. The self-hosted edition is the real product: integrations are not held back for paying customers.

• Write an integration

  Integrations are modular adapters with a documented interface, typed schemas, tests, and a template. Missing a provider you rely on? Add it, and everyone gets it.

• Or let us host it

  A managed version is planned for people who would rather not run infrastructure: OAuth apps, webhooks, sync jobs, updates, and backups handled for you.

1. Phase 1

   First release

   One timeline and one alert across your stack, with stats you choose to share and an API and MCP server to read it all.

   • Provider connections through OAuth, no pasted keys
   • Timeline of deploys, stars, signups, logins, payments, and errors
   • Alerts by email, Slack, Discord, and webhook
   • Daily and weekly digests on the same channels
   • Percentage and dollar thresholds
   • Unusual daily spend and usage flagged with no setup, and a mute per metric
   • Publish chosen metrics, one at a time
   • Opt-in public project pages
   • README badges and shareable milestone cards
   • A card of your year so far
   • Read-only REST API with scoped keys
   • Read-only MCP server for AI assistants, on the same keys
   • Allowances, remaining quota, and reset dates
   • Spend where the provider reports it
   • Spend and quota projected to the end of the period, labeled as estimates
   • Usage history
   • Deploys, payments, signups, stars, and alerts marked on the trend charts
   • Revenue against spend, signups, cost per signup, and deploys for the week or month so far
2. Phase 2

   Straight after

   Connect more of your stack, run it yourself, and share it with a team.

   • More integrations, starting with ElevenLabs
   • Open source release with self-hosting on Docker
   • Team workspaces with owner, admin, and viewer roles
   • Hide the event types you do not care about
3. Phase 3

   Later

   Ideas we will build only if users ask for them.

   • Cost optimization suggestions
   • Many more integrations