Release v1.0.2

Added

• Strict Scope Enforcement for ALLOWED_DATABASES:
  • db_list_connections: Only lists permitted databases and host servers that contain permitted databases. Sanitizes defaultDatabase and active connection pools so unallowed databases are never exposed. Adds filteredMode and allowedDatabases metadata to tool responses.
  • db_discover_databases: Filters discovered physical databases from host servers to strictly return permitted databases only.
  • db_sync_databases & npm run sync: Crawler strictly syncs and adds allowed databases; non-allowed database entries on disk are preserved safely to prevent accidental configuration data loss.
  • resolveTarget: Immediately rejects queries or connection attempts against restricted databases (Access denied). Sanitizes error messages so unallowed sub-connections are never leaked.
  • db_register_sub_connection: Prevents dynamic runtime registration of restricted databases when running in filtered mode.
  • Flexible ALLOWED_DATABASES parsing: Supports CLI flags (--databases, --database, --dbs, --db, --filter), environment variable formats (comma, semicolon, whitespace, JSON array ["db1", "db2"]), and handles quoted values cleanly.
• Auto-Populate Bare Main Connections:
  • If a main connection has no pre-configured sub-connections, permitted databases are automatically mapped to that host.
• Dedicated CLI Runner for db_list_connections:
  • Added npm run list (tsx src/list.ts) script to inspect connections and test ALLOWED_DATABASES filter rules from the terminal with human-friendly and --json outputs.

Changed

• Sanitized Parameter Descriptions & Tests:
  • Replaced internal/local database names with generic placeholders (app_db, postgres_main, mysql_main) across tool schema parameter descriptions and unit test fixtures.

---

Full Changelog: v1.0.1...v1.0.2