# Connecting your data sources

Use **Configure → Data Sources** in your project to inspect the current
provider catalog. Each entry describes its authentication, required fields,
scopes, supported metrics, synchronization behavior, and execution
capabilities. This catalog is the source of truth for what your deployment
can connect and do.

## Check support before connecting

- **Available** means a supported implementation exists for the capabilities
  shown. You may still need provider credentials, account features, resource
  selection, or provider approval.
- **Guided** means setup still depends on deployment configuration, an
  external execution bridge, or a manual step.
- **Planned** describes intended support and is not a working connection.

Examples include TractionScout Site Tracking, Plausible and PostHog for
product analytics; Stripe for revenue; and GitHub, Vercel, Cloudflare, and
Sentry for code and web operations. Google Analytics 4 and Google Search
Console may require deployment-owned OAuth configuration. Confirm each
entry's current status rather than assuming a provider logo promises all of
its features.

## Connect and verify

1. Choose the provider that supplies evidence for your current question.
2. Read the setup panel's prerequisites and requested scopes. Use an account
   authorized for the intended project, site, property, or other resource.
3. Complete the provider's OAuth flow or enter the requested credential in
   the connection form. For example, PostHog uses a scoped personal API key;
   Stripe can use a restricted read key for revenue evidence.
4. Select the intended resource when the connector offers resource discovery.
5. Run the available readiness or diagnostics check, then synchronize data.
   Inspect freshness, historical coverage, and any partial failures before
   treating the result as complete.

Reading analytics does not grant permission to publish, change budgets, or
modify provider resources. Review those capabilities separately when setting
up execution. Provider subscriptions, API access, and usage charges may also
be prerequisites.

For website collection, follow the Site Tracking setup shown for your
project and review **Configure → Privacy & tracking**. For automated
discovery of the same connection capabilities, see
[Agent CLI and MCP](/docs/agent-cli.html). Return to
[Getting started](/docs/getting-started.html) to turn connected evidence into
a first reviewable action.