Speed Up Tool::Schema Validation by 5x to 100x

## Motivation and Context

`MCP::Tool::Schema` previously used the pure-Ruby `json-schema` gem for
both construction-time metaschema validation and runtime argument /
result validation. For deep schemas this cost ~32-100ms per
construction (see issue modelcontextprotocol#364).

`json_schemer` (https://rubygems.org/gems/json_schemer, v2.5.0) is
an actively maintained alternative that is much faster on this workload.
Switching reduces cold-start warming pressure in horizontally scaled
multi-tenant deployments (the scenario raised in modelcontextprotocol#364) and brings
runtime data validation down to sub-millisecond per call once
the per-instance schemer is memoized.

The implementation preserves the legacy behavior of the `json-schema`
based code. Metaschema is pinned to draft-04 (`meta_schema:` option);
the `$schema` dialect URI is still emitted in `to_h` as 2020-12
but runtime validation continues to use draft-04, exactly as before.
Moving the runtime validator itself to 2020-12 is out of scope for
this PR and warrants its own discussion. `format` keywords are not enforced
(`format: false` option), matching what `json-schema`'s draft-04 path did.
Malformed schemas (e.g. an invalid `pattern` regex) continue to
surface as `ArgumentError, "Invalid JSON Schema: ..."`;
the `RegexpError` `json_schemer` raises during construction is wrapped.
`Symbol` values in argument data are coerced to strings inside
the internal `stringify` helper so they validate against `type: "string"`
the same way they did before.

The public `Schema#schema` accessor is dropped. It was a refactor
artifact from modelcontextprotocol#198 (the consolidation of `InputSchema`'s
`properties` / `required` readers into a single hash) with no callers
in `lib/`; tests that needed the merged hash now read it through
`to_h`. The accessor's only remaining role would have been to expose
a mutation path that could desynchronize the memoized `@schemer`,
which this change removes.

Runtime dependency delta: drops `json-schema` (and `addressable`),
adds `json_schemer`, `hana`, `regexp_parser`, and `simpleidn`. Each added
gem is a single-purpose library tied to JSON Schema spec compliance.

Closes modelcontextprotocol#364.

## How Has This Been Tested?

Side-by-side benchmark on representative schemas (simple, with `$ref` /
`additionalProperties`, nested, depth 20, depth 40):
construction-time metaschema validation is 4.7x to ~100x faster;
runtime data validation with the memoized schemer is sub-millisecond
across all sizes.

Four existing tests had their stub targets or message assertions updated
to track the new validator while preserving the original test intent.
The "unexpected errors bubble up" tests in `input_schema_test.rb`
and `output_schema_test.rb` now stub `JSONSchemer::Schema#validate`
instead of `JSON::Validator.fully_validate`. The cache tests in
`schema_test.rb` now stub `JSONSchemer::Schema#validate_schema`.
The two "detailed error message" tests in `tool_test.rb` switched from
asserting `json-schema`'s exact wording to format-agnostic substrings
(`"properties/count/minimum"`, `"number"`). The "required arguments are
converted to strings" test in `input_schema_test.rb` now reads the result
through `to_h[:required]` instead of the removed `schema` accessor.

Regression tests added for the legacy-behavior preservations above:
`format` keyword is not enforced, invalid `pattern` raises `ArgumentError`
(not `RegexpError`), and `Symbol` values validate against `type: "string"`.

## Breaking Changes

`Schema#schema` is no longer a public method. The accessor had no callers
outside the test suite, but anyone reading the internal representation
directly (rather than through `to_h`) will need to switch.
`Schema` / `InputSchema` / `OutputSchema` otherwise keep
their constructor signatures, `to_h` output, `ValidationError` /
`ArgumentError` raise points, and the `$schema` dialect emission.
The validator error message wording inside `ArgumentError` and
`ValidationError` now comes from `json_schemer` and still includes
the JSON pointer path and a description of the mismatch.

Author: koic

lib/mcp/tool/input_schema.rb

@@ -12,9 +12,9 @@ def missing_required_arguments?(arguments)
       end
 
       def missing_required_arguments(arguments)
-        return [] unless schema[:required].is_a?(Array)
+        return [] unless @schema[:required].is_a?(Array)
 
-        (schema[:required] - arguments.keys.map(&:to_s))
+        (@schema[:required] - arguments.keys.map(&:to_s))
       end
 
       def validate_arguments(arguments)

lib/mcp/tool/schema.rb

@@ -1,7 +1,7 @@
 # frozen_string_literal: true
 
 require "digest"
-require "json-schema"
+require "json_schemer"
 
 module MCP
   class Tool
@@ -38,11 +38,10 @@ def clear
 
       # JSON Schema 2020-12 is the default dialect for MCP schema definitions
       # per MCP 2025-11-25 (SEP-1613). Note: emission only — runtime validation
-      # is still performed against the JSON Schema draft-04 metaschema because
-      # the `json-schema` gem does not yet support 2020-12.
+      # is still performed against the JSON Schema draft-04 metaschema.
       JSON_SCHEMA_2020_12_URI = "https://json-schema.org/draft/2020-12/schema"
 
-      attr_reader :schema
+      DRAFT4_META_SCHEMA_URI = "http://json-schema.org/draft-04/schema#"
 
       def initialize(schema = {})
         @schema = JSON.parse(JSON.dump(schema), symbolize_names: true)
@@ -51,7 +50,7 @@ def initialize(schema = {})
       end
 
       def ==(other)
-        other.is_a?(self.class) && schema == other.schema
+        other.is_a?(self.class) && @schema == other.instance_variable_get(:@schema)
       end
 
       def to_h
@@ -62,8 +61,38 @@ def to_h
 
       private
 
+      def stringify(obj)
+        case obj
+        when Hash
+          obj.each_with_object({}) { |(k, v), h| h[k.to_s] = stringify(v) }
+        when Array
+          obj.map { |v| stringify(v) }
+        when Symbol
+          obj.to_s
+        else
+          obj
+        end
+      end
+
+      # Lazily built so a cache hit in `validate_schema!` avoids the schemer construction cost.
+      # Memoized per Schema instance because schema content is fixed at construction,
+      # so the compiled schemer is reusable across many `fully_validate` calls.
+      #
+      # `format: false` preserves the legacy behavior of the previous `json-schema` based implementation,
+      # which did not enforce `format` keywords. `RegexpError` from a malformed `pattern` is re-raised as
+      # `ArgumentError` so callers see the same exception class they used to.
+      def schemer
+        @schemer ||= JSONSchemer.schema(
+          stringify(schema_for_validation),
+          meta_schema: DRAFT4_META_SCHEMA_URI,
+          format: false,
+        )
+      rescue RegexpError => e
+        raise ArgumentError, "Invalid JSON Schema: #{e.message}"
+      end
+
       def fully_validate(data)
-        JSON::Validator.fully_validate(schema_for_validation, data)
+        schemer.validate(stringify(data)).map { |validation_error| validation_error.fetch("error") }
       end
 
       def validate_schema!
@@ -75,26 +104,16 @@ def validate_schema!
         key = Digest::SHA256.hexdigest(JSON.generate(target, max_nesting: false))
         return if VALIDATION_CACHE.validated?(key)
 
-        gem_path = File.realpath(Gem.loaded_specs["json-schema"].full_gem_path)
-        schema_reader = JSON::Schema::Reader.new(
-          accept_uri: false,
-          accept_file: ->(path) { File.realpath(path.to_s).start_with?(gem_path) },
-        )
-        metaschema_path = Pathname.new(JSON::Validator.validator_for_name("draft4").metaschema)
-        # Converts metaschema to a file URI for cross-platform compatibility
-        metaschema_uri = JSON::Util::URI.file_uri(metaschema_path.expand_path.cleanpath.to_s.tr("\\", "/"))
-        metaschema = metaschema_uri.to_s
-        errors = JSON::Validator.fully_validate(metaschema, target, schema_reader: schema_reader)
+        errors = schemer.validate_schema.map { |validation_error| validation_error.fetch("error") }
         if errors.any?
           raise ArgumentError, "Invalid JSON Schema: #{errors.join(", ")}"
         end
 
         VALIDATION_CACHE.store(key)
       end
 
-      # The `json-schema` gem's draft-04 validator cannot resolve newer or unknown `$schema`
-      # dialect URIs. Strip the top-level `$schema` before validation so a dialect URI
-      # (whether SDK-injected by `to_h` or user-supplied) does not break the validator.
+      # `json_schemer` is pinned to the draft-04 metaschema, so strip top-level `$schema` before validation:
+      # this preserves the legacy behavior of ignoring the advertised dialect URI when the SDK validates schemas.
       def schema_for_validation
         return @schema unless @schema.key?(:"$schema")
 

mcp.gemspec

@@ -30,5 +30,5 @@ Gem::Specification.new do |spec|
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
   spec.require_paths = ["lib"]
 
-  spec.add_dependency("json-schema", ">= 4.1")
+  spec.add_dependency("json_schemer", ">= 2.4")
 end

test/mcp/tool/input_schema_test.rb

@@ -7,7 +7,7 @@ class Tool
     class InputSchemaTest < ActiveSupport::TestCase
       test "required arguments are converted to strings" do
         input_schema = InputSchema.new(properties: { message: { type: "string" } }, required: [:message])
-        assert_equal ["message"], input_schema.schema[:required]
+        assert_equal ["message"], input_schema.to_h[:required]
       end
 
       test "to_h returns a hash representation of the input schema" do
@@ -139,10 +139,10 @@ class InputSchemaTest < ActiveSupport::TestCase
       test "unexpected errors bubble up from validate_arguments" do
         schema = InputSchema.new(properties: { foo: { type: "string" } }, required: ["foo"])
 
-        JSON::Validator.stub(:fully_validate, ->(*) { raise "unexpected error" }) do
-          assert_raises(RuntimeError) do
-            schema.validate_arguments({ foo: "bar" })
-          end
+        JSONSchemer::Schema.any_instance.stubs(:validate).raises("unexpected error")
+
+        assert_raises(RuntimeError) do
+          schema.validate_arguments(foo: "bar")
         end
       end
 
@@ -200,6 +200,26 @@ class InputSchemaTest < ActiveSupport::TestCase
         schema6 = InputSchema.new(properties: { foo: { type: "string" } }, required: ["foo"], additionalProperties: false)
         refute_equal schema1, schema6
       end
+
+      test "format keyword is not enforced (legacy behavior)" do
+        schema = InputSchema.new(
+          properties: { email: { type: "string", format: "email" } },
+          required: ["email"],
+        )
+        assert_nil(schema.validate_arguments(email: "not_an_email"))
+      end
+
+      test "invalid pattern raises ArgumentError, not RegexpError" do
+        error = assert_raises(ArgumentError) do
+          InputSchema.new(properties: { id: { type: "string", pattern: "[" } })
+        end
+        assert_includes error.message, "Invalid JSON Schema"
+      end
+
+      test "Symbol values in arguments are treated as strings" do
+        schema = InputSchema.new(properties: { foo: { type: "string" } }, required: ["foo"])
+        assert_nil(schema.validate_arguments(foo: :bar))
+      end
     end
   end
 end

test/mcp/tool/output_schema_test.rb

@@ -110,10 +110,10 @@ class OutputSchemaTest < ActiveSupport::TestCase
       test "unexpected errors bubble up from validate_result" do
         schema = OutputSchema.new(properties: { foo: { type: "string" } }, required: ["foo"])
 
-        JSON::Validator.stub(:fully_validate, ->(*) { raise "unexpected error" }) do
-          assert_raises(RuntimeError) do
-            schema.validate_result({ foo: "bar" })
-          end
+        JSONSchemer::Schema.any_instance.stubs(:validate).raises("unexpected error")
+
+        assert_raises(RuntimeError) do
+          schema.validate_result(foo: "bar")
         end
       end
 

test/mcp/tool/schema_test.rb

@@ -10,15 +10,15 @@ class SchemaTest < ActiveSupport::TestCase
       end
 
       test "validates a schema once and reuses the result for identical schemas" do
-        JSON::Validator.expects(:fully_validate).once.returns([])
+        JSONSchemer::Schema.any_instance.expects(:validate_schema).once.returns([])
 
         schema = { properties: { validates_once: { type: "string" } } }
         InputSchema.new(schema)
         InputSchema.new(schema)
       end
 
       test "validates distinct schemas separately" do
-        JSON::Validator.expects(:fully_validate).twice.returns([])
+        JSONSchemer::Schema.any_instance.expects(:validate_schema).twice.returns([])
 
         InputSchema.new(properties: { distinct_a: { type: "string" } })
         InputSchema.new(properties: { distinct_b: { type: "string" } })
@@ -64,11 +64,11 @@ class SchemaTest < ActiveSupport::TestCase
           break
         end
 
-        JSON::Validator.stub(:fully_validate, []) do
-          assert_nothing_raised do
-            InputSchema.new(schema)
-            InputSchema.new(schema)
-          end
+        JSONSchemer::Schema.any_instance.stubs(:validate_schema).returns([])
+
+        assert_nothing_raised do
+          InputSchema.new(schema)
+          InputSchema.new(schema)
         end
       end
 

test/mcp/tool_test.rb

@@ -167,8 +167,8 @@ class InputSchemaTool < Tool
       end
 
       assert_includes error.message, "Invalid JSON Schema"
-      assert_includes error.message, "#/properties/count/minimum"
-      assert_includes error.message, "string did not match the following type: number"
+      assert_includes error.message, "properties/count/minimum"
+      assert_includes error.message, "number"
     end
 
     test ".define allows definition of simple tools with a block" do
@@ -431,8 +431,8 @@ class OutputSchemaObjectTool < Tool
       end
 
       assert_includes error.message, "Invalid JSON Schema"
-      assert_includes error.message, "#/properties/count/minimum"
-      assert_includes error.message, "string did not match the following type: number"
+      assert_includes error.message, "properties/count/minimum"
+      assert_includes error.message, "number"
     end
 
     test "output_schema accepts $ref in schema" do