Skip to main content
Connected mode turns local attribution into shared platform context. It is optional and reversible.

​
Why connect

The CLI can calculate attribution for one developer or repository. The Autter platform can join that attribution with your software delivery workflow. Connected data supports:
  • AI adoption and acceptance metrics by contributor, team, repository, agent, and model
  • Pull request and production-level durability analysis
  • Prompt and agent-session search for review or debugging
  • Token usage and cost analysis
  • Organization audit logs for CLI access and data pushes
  • Attribution preservation across server-side squash and rebase merges when the required SCM integration is installed
The platform helps teams answer which agent workflows produce durable code. It does not change how developers prompt, stage, or commit.

​
Connect with onboarding

1

Start onboarding

Run the primary connection flow:
autter onboard
2

Choose connected mode

Select the option to connect to the Autter platform. Autter opens a browser, displays a one-time code, and waits for you to authorize the device.
3

Confirm the active identity

After onboarding finishes, run:
autter whoami
Run autter onboard --connect to choose connected mode directly. Use autter onboard --force when you previously completed onboarding and want to change your choice.
If browser authorization fails, Autter keeps the machine in local-only mode. Your local attribution continues working. Retry later with autter onboard --connect or use the PAT fallback.

​
PAT fallback

Use a personal access token only if autter onboard cannot complete the browser connection.
1

Open the dashboard

Go to app.autter.dev and sign in.
2

Create a personal access token

Open the organization you want to connect. Go to Org Settings → Access Tokens, create a token, and copy it.
3

Complete CLI sign-in

Pass the token to the CLI:
autter login --token autter_pat_xxxxxxxx
Treat the PAT like a password. Do not commit it or paste it into tickets, chat, or shared logs.
Placeholder for an Autter platform dashboard with AI authorship metrics

​
How connected storage works

An access token belongs to your account. If you belong to several organizations, Autter uses the repository’s Git remote to identify its owning organization. If no match exists, it uses your default organization. Connected mode keeps line-level attribution in local Git notes under refs/notes/ai. Prompt transcripts and usage data also sync to your organization’s Autter environment. After a commit, Autter reads any captured agent transcript, normalizes it, and redacts detected secrets before it leaves the machine. The transcript is queued as a content-addressed object. Its cas:<hash> reference is added to the matching prompt record so review and blame tools can resolve the conversation behind an attributed change. Transcript capture is best-effort. A missing, empty, or unreadable transcript does not block the commit or invalidate its authorship note. The background service retries queued uploads. Usage metrics route to the organization database associated with the signed identity and repository remote.

​
Manage access

Open Org Settings → Access Tokens in the dashboard to:
  • Review active tokens
  • Revoke a token
  • Review sign-in and data-push activity
To remove credentials from the current machine: 
autter logout
To return to local-only mode: 
autter onboard --local --force

Review data and opt-out controls

See exactly what remains local, enters Git, or uploads in connected mode.