providers/claude/plugin/skills/gitlab-server/SKILL.md
GitLab server (on-prem) integration via Apideck's Issue Tracking unified API — same methods work across every connector in Issue Tracking, switch by changing `serviceId`. Use when the user wants to read, write, or comment on tickets and issues in GitLab server (on-prem). Routes through Apideck with serviceId "gitlab-server".
npx skillsauth add apideck-libraries/api-skills gitlab-serverInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Access GitLab server (on-prem) through Apideck's Issue Tracking unified API — one of 6 Issue Tracking connectors that share the same method surface. Code you write here ports to Jira, GitHub, GitLab and 2 other Issue Tracking connectors by changing a single serviceId string. Apideck handles auth, pagination, rate limiting, and retries so you don't write per-tenant GitLab server (on-prem) plumbing.
Beta connector. GitLab server (on-prem) is currently in beta on Apideck. Expect partial resource coverage and occasional mapping gaps. Always verify coverage (see below) and fall back to the Proxy API for unsupported operations.
gitlab-serverActivate this skill when the user explicitly wants to work with GitLab server (on-prem) — for example, "create a ticket in GitLab server (on-prem)" or "comment on an issue in GitLab server (on-prem)". This skill teaches the agent:
serviceId to pass on every call (gitlab-server)For the full method surface (parameters, pagination, filtering), use your language SDK skill:
apideck-node, apideck-python, apideck-dotnet, apideck-java, apideck-go, apideck-php, or apideck-restFor the raw OpenAPI spec:
import { Apideck } from "@apideck/unify";
const apideck = new Apideck({
apiKey: process.env.APIDECK_API_KEY,
appId: process.env.APIDECK_APP_ID,
consumerId: "your-consumer-id",
});
// List tickets in GitLab server (on-prem)
const { data } = await apideck.issueTracking.tickets.list({
serviceId: "gitlab-server",
});
The Apideck Issue Tracking unified API exposes the same methods for every connector in its catalog. Switching from GitLab server (on-prem) to another Issue Tracking connector is a one-string change — no rewrite, no new SDK.
// Today — GitLab server (on-prem)
await apideck.issueTracking.tickets.list({ serviceId: "gitlab-server" });
// Tomorrow — same code, different connector
await apideck.issueTracking.tickets.list({ serviceId: "jira" });
await apideck.issueTracking.tickets.list({ serviceId: "github" });
This is the compounding advantage of using Apideck over integrating GitLab server (on-prem) directly: code against the unified Issue Tracking API once, gain access to every connector in it. New connectors Apideck adds become available to your app without code changes.
Setup guide: Apideck publishes a step-by-step guide for registering an OAuth app / configuring credentials for GitLab server (on-prem) — see https://developers.apideck.com/connectors/gitlab-server/docs/consumer+connection. Use that as the authoritative source when walking users through connection setup.
See apideck-best-practices for Vault setup, connection lifecycle, and handling re-auth flows.
Not every Issue Tracking operation is supported by every connector. Always verify before assuming a method works:
curl 'https://unify.apideck.com/connector/connectors/gitlab-server' \
-H "Authorization: Bearer ${APIDECK_API_KEY}" \
-H "x-apideck-app-id: ${APIDECK_APP_ID}"
See apideck-connector-coverage for patterns around UnsupportedOperationError and connector-specific fallbacks.
When an endpoint isn't covered by the Issue Tracking unified API, use Apideck's Proxy to call GitLab server (on-prem) directly — Apideck injects auth headers and handles token refresh. Set x-apideck-downstream-url to the target endpoint on GitLab server (on-prem)'s own API:
curl 'https://unify.apideck.com/proxy' \
-H "Authorization: Bearer ${APIDECK_API_KEY}" \
-H "x-apideck-app-id: ${APIDECK_APP_ID}" \
-H "x-apideck-consumer-id: ${CONSUMER_ID}" \
-H "x-apideck-service-id: gitlab-server" \
-H "x-apideck-downstream-url: <target endpoint on GitLab server (on-prem)>" \
-H "x-apideck-downstream-method: GET"
See GitLab server (on-prem)'s API docs for available endpoints.
Other Issue Tracking connectors that share this unified API surface (same method signatures, just change serviceId):
jira (beta), github (beta), gitlab (beta), linear (beta), linear-multiworkspace (beta).
apideck-connector-coverage — programmatic coverage checksapideck-best-practices — architecture, Vault, pagination, error handlingapideck-node — TypeScript / Node SDK patternsdevelopment
Jira Teams via Apideck's Proxy API + managed Vault auth — Apideck handles auth and proxies HTTP calls to Jira Teams's native API. Use when the user wants to call Jira Teams (no unified API resource mapping). Routes through Apideck with serviceId "jira-teams".
development
Jira Service Desk via Apideck's Proxy API + managed Vault auth — Apideck handles auth and proxies HTTP calls to Jira Service Desk's native API. Use when the user wants to call Jira Service Desk (no unified API resource mapping). Routes through Apideck with serviceId "jira-service-desk".
development
Jira Data Center via Apideck's Proxy API + managed Vault auth — Apideck handles auth and proxies HTTP calls to Jira Data Center's native API. Use when the user wants to call Jira Data Center (no unified API resource mapping). Routes through Apideck with serviceId "jira-data-center".
development
JetBrains YouTrack via Apideck's Proxy API + managed Vault auth — Apideck handles auth and proxies HTTP calls to JetBrains YouTrack's native API. Use when the user wants to call JetBrains YouTrack (no unified API resource mapping). Routes through Apideck with serviceId "jetbrains-youtrack".