Skip to content

GitLab

Liddr reads projects, issues, and merge requests the token can see during ingest, and — for reverse sync — pushes branches and opens merge requests when a Forge build completes. Both use a GitLab access token (sent as PRIVATE-TOKEN for reads; git pushes use the oauth2 HTTPS scheme).

What Liddr accesses

DirectionAccess
Ingest (read into your knowledge graph)Read-only
Reverse sync (Liddr writing back)Read & write — only when a Forge action publishes back to this source

Setup

  1. In GitLab: Preferences → Access tokens (or a Group/Project access token under the group/project's Settings → Access tokens — both work and scope access tighter than a personal token).
  2. Grant the scopes listed below and set an expiry.
  3. In Liddr: Settings → Connections → GitLab, paste the token. For a self-managed instance, set the base URL to https://<your-gitlab>/api/v4 (default is gitlab.com).
  4. Optionally set Group in the connector's Advanced settings to scope ingest to one group's projects, then Test and run an Ingest.

Required scopes

Personal / Group / Project access token

Ingest — minimum

  • read_api (covers projects, issues, and merge-request reads)
  • Token's user needs at least Reporter on the projects to ingest

Reverse sync — additionally

  • api (create / close merge requests — read_api is read-only)
  • write_repository (push build branches over HTTPS)
  • Token's user needs the Developer role on target projects

TIP

Ingest-only tokens should stop at read_api. For reverse sync, api + write_repository is also the maximum — never grant sudo, admin_mode, or runner scopes.

Reference

Liddr — grounded in your reality, linked to the evidence.