feat: upgrade mcp-trino v0.4.0 / mcp-datahub v0.6.0 with description overrides (#101)

* feat: upgrade mcp-trino v0.4.0 / mcp-datahub v0.6.0 with config-driven description overrides

Bump mcp-trino v0.3.0 → v0.4.0 and mcp-datahub v0.5.2 → v0.6.0 to pick up
improved default tool descriptions. Wire the new WithDescriptions toolkit option
through YAML config so deployers can customize tool descriptions per-instance
without code changes.

* docs: add descriptions config option to toolkit reference

Author: cjimti

configs/platform.yaml

@@ -150,6 +150,9 @@ toolkits:
       default_limit: 1000
       max_limit: 10000
       read_only: true
+      # descriptions:
+      #   trino_query: "Execute a SQL query against the data warehouse"
+      #   trino_describe_table: "Get table schema with column types and sample data"
 
   datahub:
     enabled: false
@@ -162,6 +165,9 @@ toolkits:
     config:
       search_limit: 50
       lineage_max_depth: 5
+      # descriptions:
+      #   datahub_search: "Search the data catalog for datasets and dashboards"
+      #   datahub_get_entity: "Get full metadata for a catalog entity by URN"
 
   s3:
     enabled: false

docs/llms-full.txt

@@ -326,6 +326,7 @@ The `database` store requires `database.dsn`. Database-backed sessions survive r
 | `default_limit` | int | 1000 | Default row limit |
 | `max_limit` | int | 10000 | Maximum row limit |
 | `read_only` | bool | false | Restrict to read-only queries |
+| `descriptions` | map | {} | Override tool descriptions (key: tool name, value: description text) |
 
 ### DataHub
 
@@ -336,6 +337,7 @@ The `database` store requires `database.dsn`. Database-backed sessions survive r
 | `timeout` | duration | 30s | API request timeout |
 | `default_limit` | int | 10 | Default search limit |
 | `max_limit` | int | 100 | Maximum search limit |
+| `descriptions` | map | {} | Override tool descriptions (key: tool name, value: description text) |
 
 ### S3
 

docs/reference/configuration.md

@@ -244,6 +244,9 @@ toolkits:
       max_limit: 10000
       read_only: false
       connection_name: "Production"
+      descriptions:
+        trino_query: "Execute SQL with automatic semantic enrichment from DataHub"
+        trino_describe_table: "Get table schema with DataHub context — the richest single-call way to understand a table"
 ```
 
 | Option | Type | Default | Description |
@@ -261,6 +264,7 @@ toolkits:
 | `max_limit` | int | `10000` | Maximum row limit |
 | `read_only` | bool | `false` | Restrict to read-only queries |
 | `connection_name` | string | instance name | Display name |
+| `descriptions` | map | `{}` | Override tool descriptions (key: tool name, value: description text) |
 
 ### DataHub
 
@@ -276,6 +280,9 @@ toolkits:
       max_lineage_depth: 5
       connection_name: "Primary Catalog"
       debug: false
+      descriptions:
+        datahub_search: "Search the data catalog for datasets and dashboards"
+        datahub_get_entity: "Get full metadata for a catalog entity by URN"
 ```
 
 | Option | Type | Default | Description |
@@ -288,6 +295,7 @@ toolkits:
 | `max_lineage_depth` | int | `5` | Maximum lineage depth |
 | `connection_name` | string | instance name | Display name |
 | `debug` | bool | `false` | Enable debug logging for GraphQL operations |
+| `descriptions` | map | `{}` | Override tool descriptions (key: tool name, value: description text) |
 
 ### S3
 

go.mod

@@ -14,9 +14,9 @@ require (
 	github.com/swaggo/swag v1.16.4
 	github.com/testcontainers/testcontainers-go v0.40.0
 	github.com/testcontainers/testcontainers-go/modules/postgres v0.40.0
-	github.com/txn2/mcp-datahub v0.5.2
+	github.com/txn2/mcp-datahub v0.6.0
 	github.com/txn2/mcp-s3 v0.1.4
-	github.com/txn2/mcp-trino v0.3.0
+	github.com/txn2/mcp-trino v0.4.0
 	golang.org/x/crypto v0.47.0
 	gopkg.in/yaml.v3 v3.0.1
 )

go.sum

@@ -264,12 +264,12 @@ github.com/tklauser/numcpus v0.6.1 h1:ng9scYS7az0Bk4OZLvrNXNSAO2Pxr1XXRAPyjhIx+F
 github.com/tklauser/numcpus v0.6.1/go.mod h1:1XfjsgE2zo8GVw7POkMbHENHzVg3GzmoZ9fESEdAacY=
 github.com/trinodb/trino-go-client v0.333.0 h1:+bsW8/uLFNF00MEL9JZJym94LlUnle25VgDlWGPEZos=
 github.com/trinodb/trino-go-client v0.333.0/go.mod h1:91okdYtRUZoj3XJu/tqdzu11sNliQuN4A+vMFEB8GVE=
-github.com/txn2/mcp-datahub v0.5.2 h1:Nq2VetunmEwyW2kR7xnpV9Wl487BFnZ+PpKSgDZjh+U=
-github.com/txn2/mcp-datahub v0.5.2/go.mod h1:9YWiNjhmFjmXMK9PX9QmlQgouaZLu/edqJ5hKC5qZUA=
+github.com/txn2/mcp-datahub v0.6.0 h1:8iiaQ9xR67Qt8B84QK6TtdljP4hFEKYLBzZQDVA1BTY=
+github.com/txn2/mcp-datahub v0.6.0/go.mod h1:UyXoTLT9H5DFkrdi/k0G82x+fZN5N4PSjom6FPGTkI0=
 github.com/txn2/mcp-s3 v0.1.4 h1:73VfwofCNUoVAzWQgh1LoaAsNW42jEUTGfRPe1s7Vkk=
 github.com/txn2/mcp-s3 v0.1.4/go.mod h1:yu2m8DZdsHJEWj+Ktkhz1XUk17jqPdVSa4UcyZWetkc=
-github.com/txn2/mcp-trino v0.3.0 h1:q3mLQPvgDDfODZNgL86JKUrHaDUxpdWhNzaOt6iu6sM=
-github.com/txn2/mcp-trino v0.3.0/go.mod h1:1lDb2HYgMtMQWjF7NcC1cRmkXsWMu2SWq85oc4qIriU=
+github.com/txn2/mcp-trino v0.4.0 h1:QrCNv0d6NR4OW3+AhIiVYX2tDyph6uDm5tRLz3/UQyk=
+github.com/txn2/mcp-trino v0.4.0/go.mod h1:6m5jlHTExGTr2swFRFp2McDBHti8l/F7mHi2b1cYHk4=
 github.com/xeipuuv/gojsonpointer v0.0.0-20190905194746-02993c407bfb h1:zGWFAtiMcyryUHoUjUJX0/lt1H2+i2Ka2n+D3DImSNo=
 github.com/xeipuuv/gojsonpointer v0.0.0-20190905194746-02993c407bfb/go.mod h1:N2zxlSyiKSe5eX1tZViRH5QA0qijqEDrYZiPEAiq3wU=
 github.com/xeipuuv/gojsonreference v0.0.0-20180127040603-bd5ef7bd5415 h1:EzJWgHovont7NscjpAxXsDA8S8BMYve8Y5+7cuRE7R0=

pkg/toolkits/datahub/config.go

@@ -42,6 +42,9 @@ func ParseConfig(cfg map[string]any) (Config, error) {
 	// Optional bool fields
 	c.Debug = getBool(cfg, "debug", false)
 
+	// Optional description overrides
+	c.Descriptions = getStringMap(cfg, "descriptions")
+
 	return c, nil
 }
 
@@ -64,6 +67,21 @@ func getInt(cfg map[string]any, key string, defaultVal int) int {
 	return defaultVal
 }
 
+// getStringMap extracts a map[string]string value from a config map.
+func getStringMap(cfg map[string]any, key string) map[string]string { //nolint:unparam // consistent with getString/getInt helpers
+	raw, ok := cfg[key].(map[string]any)
+	if !ok {
+		return nil
+	}
+	result := make(map[string]string, len(raw))
+	for k, v := range raw {
+		if s, ok := v.(string); ok {
+			result[k] = s
+		}
+	}
+	return result
+}
+
 // getDuration extracts a duration value from a config map.
 func getDuration(cfg map[string]any, key string) (time.Duration, error) {
 	if v, ok := cfg[key].(string); ok {

pkg/toolkits/datahub/config_test.go

@@ -231,6 +231,93 @@ func TestDatahubGetInt(t *testing.T) {
 	}
 }
 
+func TestGetStringMap(t *testing.T) {
+	t.Run("valid map", func(t *testing.T) {
+		cfg := map[string]any{
+			"descriptions": map[string]any{
+				"datahub_search":     "Search the catalog",
+				"datahub_get_entity": "Get entity details",
+			},
+		}
+		result := getStringMap(cfg, "descriptions")
+		if len(result) != 2 {
+			t.Fatalf("expected 2 entries, got %d", len(result))
+		}
+		if result["datahub_search"] != "Search the catalog" {
+			t.Errorf("datahub_search = %q", result["datahub_search"])
+		}
+		if result["datahub_get_entity"] != "Get entity details" {
+			t.Errorf("datahub_get_entity = %q", result["datahub_get_entity"])
+		}
+	})
+
+	t.Run("missing key", func(t *testing.T) {
+		cfg := map[string]any{}
+		result := getStringMap(cfg, "descriptions")
+		if result != nil {
+			t.Errorf("expected nil for missing key, got %v", result)
+		}
+	})
+
+	t.Run("wrong type", func(t *testing.T) {
+		cfg := map[string]any{"descriptions": "not a map"}
+		result := getStringMap(cfg, "descriptions")
+		if result != nil {
+			t.Errorf("expected nil for wrong type, got %v", result)
+		}
+	})
+
+	t.Run("skips non-string values", func(t *testing.T) {
+		cfg := map[string]any{
+			"descriptions": map[string]any{
+				"valid":   "a string",
+				"invalid": dhCfgTestNumVal,
+			},
+		}
+		result := getStringMap(cfg, "descriptions")
+		if len(result) != 1 {
+			t.Fatalf("expected 1 entry (non-string skipped), got %d", len(result))
+		}
+		if result["valid"] != "a string" {
+			t.Errorf("valid = %q", result["valid"])
+		}
+	})
+}
+
+func TestParseConfig_WithDescriptions(t *testing.T) {
+	cfg := map[string]any{
+		dhCfgTestURLKey: dhCfgTestExampleURL,
+		"descriptions": map[string]any{
+			"datahub_search": "Custom search description",
+		},
+	}
+
+	result, err := ParseConfig(cfg)
+	if err != nil {
+		t.Fatalf(dhCfgTestUnexpectedErr, err)
+	}
+	if len(result.Descriptions) != 1 {
+		t.Fatalf("expected 1 description, got %d", len(result.Descriptions))
+	}
+	if result.Descriptions["datahub_search"] != "Custom search description" {
+		t.Errorf("datahub_search description = %q", result.Descriptions["datahub_search"])
+	}
+}
+
+func TestParseConfig_NoDescriptions(t *testing.T) {
+	cfg := map[string]any{
+		dhCfgTestURLKey: dhCfgTestExampleURL,
+	}
+
+	result, err := ParseConfig(cfg)
+	if err != nil {
+		t.Fatalf(dhCfgTestUnexpectedErr, err)
+	}
+	if result.Descriptions != nil {
+		t.Errorf("expected nil descriptions, got %v", result.Descriptions)
+	}
+}
+
 func TestDatahubGetDuration(t *testing.T) {
 	cfg := map[string]any{
 		dhCfgTestString:  "5m",

pkg/toolkits/datahub/toolkit.go

@@ -29,14 +29,15 @@ const (
 
 // Config holds DataHub toolkit configuration.
 type Config struct {
-	URL             string        `yaml:"url"`
-	Token           string        `yaml:"token"`
-	Timeout         time.Duration `yaml:"timeout"`
-	DefaultLimit    int           `yaml:"default_limit"`
-	MaxLimit        int           `yaml:"max_limit"`
-	MaxLineageDepth int           `yaml:"max_lineage_depth"`
-	ConnectionName  string        `yaml:"connection_name"`
-	Debug           bool          `yaml:"debug"` // Enable debug logging
+	URL             string            `yaml:"url"`
+	Token           string            `yaml:"token"`
+	Timeout         time.Duration     `yaml:"timeout"`
+	DefaultLimit    int               `yaml:"default_limit"`
+	MaxLimit        int               `yaml:"max_limit"`
+	MaxLineageDepth int               `yaml:"max_lineage_depth"`
+	ConnectionName  string            `yaml:"connection_name"`
+	Debug           bool              `yaml:"debug"` // Enable debug logging
+	Descriptions    map[string]string `yaml:"descriptions"`
 }
 
 // Toolkit wraps mcp-datahub toolkit for the platform.
@@ -119,14 +120,30 @@ func createClient(cfg Config) (*dhclient.Client, error) {
 	return client, nil
 }
 
+// toDataHubToolNames converts a generic string map to typed ToolName keys.
+func toDataHubToolNames(m map[string]string) map[dhtools.ToolName]string {
+	if m == nil {
+		return nil
+	}
+	result := make(map[dhtools.ToolName]string, len(m))
+	for k, v := range m {
+		result[dhtools.ToolName(k)] = v
+	}
+	return result
+}
+
 // createToolkit creates the mcp-datahub toolkit.
 func createToolkit(client *dhclient.Client, cfg Config) *dhtools.Toolkit {
+	var opts []dhtools.ToolkitOption
+	if len(cfg.Descriptions) > 0 {
+		opts = append(opts, dhtools.WithDescriptions(toDataHubToolNames(cfg.Descriptions)))
+	}
 	return dhtools.NewToolkit(client, dhtools.Config{
 		DefaultLimit:    cfg.DefaultLimit,
 		MaxLimit:        cfg.MaxLimit,
 		MaxLineageDepth: cfg.MaxLineageDepth,
 		Debug:           cfg.Debug,
-	})
+	}, opts...)
 }
 
 // Kind returns the toolkit kind.

pkg/toolkits/datahub/toolkit_test.go

@@ -6,6 +6,7 @@ import (
 	"time"
 
 	"github.com/modelcontextprotocol/go-sdk/mcp"
+	dhtools "github.com/txn2/mcp-datahub/pkg/tools"
 
 	"github.com/txn2/mcp-data-platform/pkg/query"
 	"github.com/txn2/mcp-data-platform/pkg/semantic"
@@ -327,6 +328,41 @@ func TestToolkit_ClientAndClose(t *testing.T) {
 	}
 }
 
+func TestToDataHubToolNames(t *testing.T) {
+	t.Run("nil input", func(t *testing.T) {
+		result := toDataHubToolNames(nil)
+		if result != nil {
+			t.Errorf("expected nil, got %v", result)
+		}
+	})
+
+	t.Run("valid conversion", func(t *testing.T) {
+		input := map[string]string{
+			"datahub_search":     "Custom search",
+			"datahub_get_entity": "Custom entity",
+		}
+		result := toDataHubToolNames(input)
+		if len(result) != 2 {
+			t.Fatalf("expected 2 entries, got %d", len(result))
+		}
+		for k, v := range input {
+			if got := result[dhtools.ToolName(k)]; got != v {
+				t.Errorf("result[%q] = %q, want %q", k, got, v)
+			}
+		}
+	})
+
+	t.Run("empty map", func(t *testing.T) {
+		result := toDataHubToolNames(map[string]string{})
+		if result == nil {
+			t.Error("expected non-nil empty map")
+		}
+		if len(result) != 0 {
+			t.Errorf("expected 0 entries, got %d", len(result))
+		}
+	})
+}
+
 func TestToolkit_RegisterTools(_ *testing.T) {
 	tk := newTestDatahubToolkit()
 	// Should not panic with nil server

pkg/toolkits/trino/config.go

@@ -46,6 +46,9 @@ func ParseConfig(cfg map[string]any) (Config, error) {
 		c.Timeout = timeout
 	}
 
+	// Optional description overrides
+	c.Descriptions = getStringMap(cfg, "descriptions")
+
 	return c, nil
 }
 
@@ -84,6 +87,21 @@ func getBoolDefault(cfg map[string]any, key string, defaultVal bool) bool {
 	return defaultVal
 }
 
+// getStringMap extracts a map[string]string value from a config map.
+func getStringMap(cfg map[string]any, key string) map[string]string { //nolint:unparam // consistent with getString/getInt helpers
+	raw, ok := cfg[key].(map[string]any)
+	if !ok {
+		return nil
+	}
+	result := make(map[string]string, len(raw))
+	for k, v := range raw {
+		if s, ok := v.(string); ok {
+			result[k] = s
+		}
+	}
+	return result
+}
+
 // getDuration extracts a duration value from a config map.
 func getDuration(cfg map[string]any, key string) (time.Duration, error) {
 	if v, ok := cfg[key].(string); ok {