Skip to content
Anubhav Rai edited this page Apr 9, 2026 · 4 revisions

Stackwatch Wiki

Welcome to the Stackwatch documentation. Stackwatch monitors usage across your developer tools and alerts you before you hit limits.

How Stackwatch Is Organized

Stackwatch uses a two-level hierarchy:

  • Organizations — a top-level grouping (e.g. a company, team, or personal workspace)
  • Projects — individual projects within an organization (e.g. prod, staging, side-project)

Each project has its own set of Integrations, Alerts, and Settings. Integrations are connected at the project level — each project independently tracks usage for whichever services you connect to it.

Navigation:

  1. The Dashboard shows all your organizations and their projects.
  2. Click into a project to reach that project's Dashboard, Integrations, Alerts, and Settings tabs.
  3. Add or manage integrations from a project's Integrations tab.

On the Free plan, you can have one primary organization and one primary project per organization. Additional organizations or projects are paused until you upgrade or make one primary. Paused orgs/projects can be deleted from the main dashboard.


Connecting Services

Each service requires an API key or token. Keys are encrypted before being stored — Stackwatch never logs or exposes them in plain text.

Connect a service from a project's Integrations tab → Add account.

Service What you need Guide
GitHub Actions Personal Access Token Connecting GitHub Actions
Vercel API Token Connecting Vercel
Supabase Management API Key + Project Ref Connecting Supabase
Railway API Token Connecting Railway
MongoDB Atlas Atlas Public Key + Private Key + Project ID Connecting MongoDB Atlas

On the Free plan, each project supports one account per service. Upgrade to Pro to connect multiple accounts per service.


Tracked Metrics by Tier

Free

  • Aggregate usage per service (Actions minutes, bandwidth, DB size, etc.)
  • Per-project / per-bucket breakdowns where applicable (Railway per-project, Supabase per-bucket)
  • Email alerts only
  • 15-minute polling
  • 7-day history

Pro

  • Everything in Free
  • Deeper per-entity breakdowns (per-repo, per-table, per-project network/disk)
  • All alert channels (Email, Slack, Discord, Browser Push)
  • 5-minute polling
  • 30-day history + usage graphs
  • Spike detection alerts — notified when a metric spikes unusually above its recent baseline

Status Indicators

Color Meaning
Green Under 60% of limit
Yellow 60–80% of limit
Red Over 80% of limit

Polling

The worker polls all connected integrations every 15 minutes (Free) or 5 minutes (Pro). The Last synced timestamp on each integration card shows when data was last fetched. If an integration is in an error state, the worker will keep retrying on the next cycle without affecting your other integrations.

Clone this wiki locally