diff --git a/src/basic_memory/schemas/memory_url.py b/src/basic_memory/schemas/memory_url.py
new file mode 100644
index 00000000..9f45ae07
--- /dev/null
+++ b/src/basic_memory/schemas/memory_url.py
@@ -0,0 +1,99 @@
+"""Memory URL schema for knowledge addressing.
+
+The memory:// URL scheme provides a unified way to address knowledge across projects:
+memory://project-name/path/to/content
+
+The host portion (project-name) identifies the knowledge base context, while
+the path represents a relative permalink within that project.
+
+Examples:
+    memory://basic-memory/specs/search/*           # Pattern matching
+    memory://basic-memory/topic/search~ranking     # Fuzzy search
+    memory://basic-memory/specs/link-resolution    # Exact permalink
+    memory://basic-memory/related/sync             # Related content
+    memory://basic-memory/context/current          # Context building
+"""
+
+from typing import Optional, Dict, Any
+from pydantic import BaseModel, Field, field_validator, ValidationError
+
+
+class MemoryUrl(BaseModel):
+    """memory:// URL scheme for knowledge addressing."""
+    
+    scheme: str = Field(default="memory", frozen=True)
+    host: str
+    path: str
+    
+    # Special fields for pattern matching and context
+    pattern: Optional[str] = None
+    fuzzy: Optional[str] = None
+    params: Dict[str, Any] = Field(default_factory=dict)
+    
+    @field_validator("scheme")
+    @classmethod
+    def validate_scheme(cls, v: str) -> str:
+        """Validate URL scheme."""
+        if v != "memory":
+            raise ValueError("URL must use memory:// scheme")
+        return v
+    
+    @field_validator("host")
+    @classmethod
+    def validate_host(cls, v: Optional[str]) -> str:
+        """Validate host (project identifier)."""
+        if not v:
+            raise ValueError("URL must include project/context identifier")
+        return v
+        
+    @classmethod
+    def parse(cls, url_str: str) -> "MemoryUrl":
+        """Parse a memory:// URL string."""
+        # Split scheme and rest
+        if "://" not in url_str:
+            raise ValueError("URL must include scheme (memory://)")
+            
+        scheme, rest = url_str.split("://", 1)
+            
+        # Split host and path
+        parts = rest.split("/", 1)
+        if len(parts) != 2:
+            raise ValueError("URL must include both host and path")
+            
+        host, path = parts
+        path = "/" + path  # Add leading slash for consistency
+        
+        # Create base URL
+        url = cls(scheme=scheme, host=host, path=path)
+        
+        # Parse special patterns
+        path_no_slash = path[1:] if path.startswith('/') else path
+        
+        # Handle glob patterns
+        if '*' in path_no_slash:
+            url.pattern = path_no_slash.replace('*', '%')
+            
+        # Handle fuzzy search
+        if '~' in path_no_slash:
+            url.fuzzy = path_no_slash.replace('~', ' ')
+            
+        # Extract special prefixes
+        segments = path_no_slash.split('/')
+        if segments and segments[0] in {'related', 'context'}:
+            url.params['type'] = segments[0]
+            url.params['target'] = '/'.join(segments[1:])
+            
+        return url
+    
+    @property
+    def project(self) -> str:
+        """Get the project/context identifier."""
+        return self.host
+        
+    def relative_path(self) -> str:
+        """Get the path without leading slash."""
+        return self.path[1:] if self.path.startswith('/') else self.path
+        
+    def __str__(self) -> str:
+        """Convert back to URL string."""
+        return f"memory://{self.host}{self.path}"
\ No newline at end of file
diff --git a/tests/schemas/test_memory_url.py b/tests/schemas/test_memory_url.py
new file mode 100644
index 00000000..1aba4c5d
--- /dev/null
+++ b/tests/schemas/test_memory_url.py
@@ -0,0 +1,103 @@
+"""Tests for MemoryUrl parsing."""
+
+import pytest
+from pydantic import ValidationError
+from basic_memory.schemas.memory_url import MemoryUrl
+
+
+def test_basic_permalink():
+    """Test basic permalink parsing."""
+    url = MemoryUrl.parse("memory://basic-memory/specs/search")
+    assert url.scheme == "memory"
+    assert url.host == "basic-memory"
+    assert url.path == "/specs/search"
+    assert url.pattern is None
+    assert url.fuzzy is None
+    assert url.params == {}
+
+
+def test_glob_pattern():
+    """Test glob pattern conversion."""
+    url = MemoryUrl.parse("memory://basic-memory/specs/search/*")
+    assert url.host == "basic-memory"
+    assert url.path == "/specs/search/*"
+    assert url.pattern == "specs/search/%"
+    assert url.fuzzy is None
+
+
+def test_fuzzy_search():
+    """Test fuzzy search term parsing."""
+    url = MemoryUrl.parse("memory://basic-memory/topic/search~ranking")
+    assert url.host == "basic-memory"
+    assert url.path == "/topic/search~ranking"
+    assert url.pattern is None
+    assert url.fuzzy == "topic/search ranking"
+
+
+def test_related_prefix():
+    """Test related content prefix."""
+    url = MemoryUrl.parse("memory://basic-memory/related/specs/search")
+    assert url.host == "basic-memory"
+    assert url.path == "/related/specs/search"
+    assert url.params == {
+        "type": "related",
+        "target": "specs/search"
+    }
+
+
+def test_context_prefix():
+    """Test context prefix."""
+    url = MemoryUrl.parse("memory://basic-memory/context/current")
+    assert url.host == "basic-memory"
+    assert url.path == "/context/current"
+    assert url.params == {
+        "type": "context",
+        "target": "current"
+    }
+
+
+def test_invalid_scheme():
+    """Test that other schemes are rejected."""
+    with pytest.raises(ValidationError):
+        MemoryUrl.parse("http://basic-memory/specs/search")
+
+
+def test_missing_host():
+    """Test that host is required."""
+    with pytest.raises(ValidationError):
+        MemoryUrl.parse("memory:///specs/search")
+
+
+def test_complex_pattern():
+    """Test multiple glob patterns."""
+    url = MemoryUrl.parse("memory://basic-memory/specs/*/search/*")
+    assert url.host == "basic-memory"
+    assert url.path == "/specs/*/search/*"
+    assert url.pattern == "specs/%/search/%"
+
+
+def test_complex_fuzzy():
+    """Test multiple fuzzy terms."""
+    url = MemoryUrl.parse("memory://basic-memory/specs/search~ranking~performance")
+    assert url.host == "basic-memory"
+    assert url.path == "/specs/search~ranking~performance"
+    assert url.fuzzy == "specs/search ranking performance"
+
+
+def test_url_reconstruction():
+    """Test converting back to string."""
+    original = "memory://basic-memory/specs/search"
+    url = MemoryUrl.parse(original)
+    assert str(url) == original
+
+
+def test_relative_path():
+    """Test getting path without leading slash."""
+    url = MemoryUrl.parse("memory://basic-memory/specs/search")
+    assert url.relative_path() == "specs/search"
+
+
+def test_project_property():
+    """Test project name access."""
+    url = MemoryUrl.parse("memory://basic-memory/specs/search")
+    assert url.project == "basic-memory"
\ No newline at end of file