Merge pull request #1026 from rudycelekli/contrib/gods-005-tool-schema-codepoints-20261011-reviewed

fix(tools): enforce string schema lengths by Unicode code point

CHANGELOG.md

@@ -24,6 +24,8 @@
   the five borough cores load and `CCTV_NYC_MAX_SOURCES` changes the cap. The catalog-wide ceiling is
   unchanged; extra cameras share it round-robin with the other packs.
 
+- Tool string length validation counts Unicode characters correctly, including characters outside the basic multilingual plane.
+
 - Director timeline seeks cross the dateline along the short longitude arc.
 
 - Treat explicitly current, complete empty vessel coverage as a successful update,

docs/CURRENT-STATE.md

@@ -73,6 +73,8 @@ Windows/Claude Desktop Cowork confirmation remains outstanding (#927).
 
 ## Tools and local MCP server — October 1, 2026
 
+Tool argument string lengths follow JSON Schema Unicode code-point counts, so supplementary characters count once rather than as two UTF-16 units.
+
 `gods-eye-view/tools` defines queries that answer questions from the app's data,
 and `gods-eye-view/tools/mcp` exposes a composed catalog over the Model Context
 Protocol. `npm run mcp` serves Core's tools over stdio to a local MCP client,

src/tools/schema.js

@@ -106,10 +106,20 @@ export function validateValue(schema, value, path = 'arguments') {
         fail('must be a string');
         break;
       }
-      if (schema.minLength != null && value.length < schema.minLength)
-        fail(`must be at least ${schema.minLength} characters`);
-      if (schema.maxLength != null && value.length > schema.maxLength)
-        fail(`must be at most ${schema.maxLength} characters`);
+      if (schema.minLength != null || schema.maxLength != null) {
+        // Code points, as JSON Schema counts. Counting stops once both bounds
+        // are decided, so a huge argument costs no more than its limit.
+        const limit = Math.max(
+          schema.minLength ?? 0,
+          schema.maxLength != null ? schema.maxLength + 1 : 0,
+        );
+        let length = 0;
+        for (const _ of value) if (++length >= limit) break;
+        if (schema.minLength != null && length < schema.minLength)
+          fail(`must be at least ${schema.minLength} characters`);
+        if (schema.maxLength != null && length > schema.maxLength)
+          fail(`must be at most ${schema.maxLength} characters`);
+      }
       if (
         schema.pattern !== undefined &&
         !new RegExp(schema.pattern, 'u').test(value)

src/tools/tool-schema-codepoints.test.mjs

@@ -0,0 +1,33 @@
+import test from 'node:test';
+import assert from 'node:assert/strict';
+import { validateValue } from './schema.js';
+test('tool string schemas count Unicode characters rather than UTF-16 units', () => {
+  const value = '𠮷野家😀';
+  assert.deepEqual(validateValue({ type: 'string', maxLength: 4 }, value), []);
+  assert.equal(
+    validateValue({ type: 'string', minLength: 5 }, value).length,
+    1,
+  );
+  assert.equal(
+    validateValue({ type: 'string', maxLength: 3 }, value).length,
+    1,
+  );
+  assert.deepEqual(
+    validateValue({ type: 'string', minLength: 4, maxLength: 4 }, 'abcd'),
+    [],
+  );
+  assert.equal(validateValue({ type: 'string', minLength: 2 }, '𠮷').length, 1);
+});
+
+test('tool string bounds hold at their edges and on huge arguments', () => {
+  const string = (bounds) => ({ type: 'string', ...bounds });
+  assert.deepEqual(validateValue(string({ maxLength: 3 }), '😀😀😀'), []);
+  assert.equal(validateValue(string({ maxLength: 3 }), '😀😀😀😀').length, 1);
+  assert.deepEqual(validateValue(string({ minLength: 3 }), '😀😀😀'), []);
+  assert.deepEqual(validateValue(string({}), 'x'.repeat(1e6)), []);
+  // Counting all of it takes roughly 0.7 s; stopping at the limit takes none.
+  const huge = 'x'.repeat(1e8);
+  const started = performance.now();
+  assert.equal(validateValue(string({ maxLength: 60 }), huge).length, 1);
+  assert.ok(performance.now() - started < 200, 'stops counting past the limit');
+});