Only show installable projects in 'databricks labs list'

NEXT_CHANGELOG.md

@@ -9,7 +9,7 @@
 * Increase the SSH server startup timeout from 10 to 45 minutes when a GPU accelerator is requested via `databricks ssh connect --accelerator` ([#5569](https://github.com/databricks/cli/pull/5569)).
 * Fix authentication falling back to the default profile in `.databrickscfg` when a host is already configured via the environment (e.g. `DATABRICKS_HOST` with `DATABRICKS_TOKEN`) ([#5616](https://github.com/databricks/cli/pull/5616)).
 * ssh: fix opening remote environment in Cursor, which previously hung on default-extension install and never opened the editor ([#5619](https://github.com/databricks/cli/pull/5619)).
-
+* `databricks labs list` now only shows projects that can be installed ([#5560](https://github.com/databricks/cli/pull/5560)).
 
 ### Bundles
 * Remove API enum values and types that are still in development from the `databricks-bundles` Python package; these were never accepted by the backend ([#5484](https://github.com/databricks/cli/pull/5484)).

cmd/labs/list.go

@@ -2,13 +2,23 @@ package labs
 
 import (
 	"context"
+	"slices"
 
 	"github.com/databricks/cli/cmd/labs/github"
 	"github.com/databricks/cli/cmd/labs/project"
 	"github.com/databricks/cli/libs/cmdio"
 	"github.com/spf13/cobra"
 )
 
+const (
+	labsOrg = "databrickslabs"
+
+	// installableTopic is the GitHub repository topic that labs maintainers add to
+	// projects installable via `databricks labs install`. The repositories API
+	// returns topics inline, so filtering on it costs no extra requests.
+	installableTopic = "databricks-cli-installable"
+)
+
 type labsMeta struct {
 	Name        string `json:"name"`
 	Description string `json:"description"`
@@ -20,14 +30,35 @@ func allRepos(ctx context.Context) (github.Repositories, error) {
 	if err != nil {
 		return nil, err
 	}
-	cache := github.NewRepositoryCache("databrickslabs", cacheDir)
+	cache := github.NewRepositoryCache(labsOrg, cacheDir)
 	return cache.Load(ctx)
 }
 
+// installableRepos returns the org repositories that `databricks labs install` can
+// install. Most repositories don't ship a labs.yml manifest (e.g. libraries
+// published to package indexes); maintainers tag the installable ones with
+// installableTopic so the listing doesn't advertise projects that fail to install.
+func installableRepos(ctx context.Context) (github.Repositories, error) {
+	repos, err := allRepos(ctx)
+	if err != nil {
+		return nil, err
+	}
+	var out github.Repositories
+	for _, repo := range repos {
+		if repo.IsArchived || repo.IsFork {
+			continue
+		}
+		if slices.Contains(repo.Topics, installableTopic) {
+			out = append(out, repo)
+		}
+	}
+	return out, nil
+}
+
 func newListCommand() *cobra.Command {
 	return &cobra.Command{
 		Use:   "list",
-		Short: "List all labs",
+		Short: "List labs that can be installed",
 		Annotations: map[string]string{
 			"template": cmdio.Heredoc(`
 			Name	Description
@@ -37,18 +68,12 @@ func newListCommand() *cobra.Command {
 		},
 		RunE: func(cmd *cobra.Command, args []string) error {
 			ctx := cmd.Context()
-			repositories, err := allRepos(ctx)
+			repositories, err := installableRepos(ctx)
 			if err != nil {
 				return err
 			}
 			var info []labsMeta
 			for _, v := range repositories {
-				if v.IsArchived {
-					continue
-				}
-				if v.IsFork {
-					continue
-				}
 				description := v.Description
 				if len(description) > 50 {
 					description = description[:50] + "..."

cmd/labs/list_test.go

@@ -1,10 +1,14 @@
 package labs_test
 
 import (
+	"net/http"
+	"net/http/httptest"
 	"testing"
 
+	"github.com/databricks/cli/cmd/labs/github"
 	"github.com/databricks/cli/internal/testcli"
 	"github.com/databricks/cli/libs/env"
+	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
 
@@ -15,4 +19,28 @@ func TestListingWorks(t *testing.T) {
 	stdout, _, err := c.Run()
 	require.NoError(t, err)
 	require.Contains(t, stdout.String(), "ucx")
+	// blueprint is in the repositories cache fixture but lacks the
+	// databricks-cli-installable topic, proving the topic filter is applied.
+	require.NotContains(t, stdout.String(), "blueprint")
+}
+
+func TestListingFiltersReposWithoutTopic(t *testing.T) {
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		assert.Equal(t, "/users/databrickslabs/repos", r.URL.Path)
+		_, err := w.Write([]byte(`[
+			{"name": "ucx", "description": "Unity Catalog Migrations", "topics": ["databricks-cli-installable"]},
+			{"name": "brickster", "description": "R interface to Databricks", "topics": []}
+		]`))
+		assert.NoError(t, err)
+	}))
+	defer server.Close()
+	ctx := t.Context()
+	ctx = github.WithApiOverride(ctx, server.URL)
+	ctx = env.WithUserHomeDir(ctx, t.TempDir())
+
+	c := testcli.NewRunner(t, ctx, "labs", "list")
+	stdout, _, err := c.Run()
+	require.NoError(t, err)
+	require.Contains(t, stdout.String(), "ucx")
+	require.NotContains(t, stdout.String(), "brickster")
 }

cmd/labs/project/testdata/installed-in-home/.databricks/labs/databrickslabs-repositories.json

@@ -25,7 +25,9 @@
       "stargazers_count": 100500,
       "fork": false,
       "archived": false,
-      "topics": [],
+      "topics": [
+        "databricks-cli-installable"
+      ],
       "html_url": "https://github.com/databrickslabs/ucx",
       "clone_url": "https://github.com/databrickslabs/ucx.git",
       "ssh_url": "git@github.com:databrickslabs/ucx.git",