yourmemory 1.2.0__tar.gz → 1.2.1__tar.gz

{yourmemory-1.2.0/yourmemory.egg-info → yourmemory-1.2.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: yourmemory
-Version: 1.2.0
+Version: 1.2.1
 Summary: Persistent memory for Claude — Ebbinghaus forgetting curve, semantic deduplication, MCP-native
 Author-email: Sachit Misra <mishrasachit1@gmail.com>
 License:                                  Apache License
@@ -181,7 +181,7 @@ Requires-Dist: numpy
 Requires-Dist: python-dateutil
 Requires-Dist: duckdb>=0.10.0
 Requires-Dist: apscheduler
-Requires-Dist: spacy<4.0,>=3.7
+Requires-Dist: spacy<4.0,>=3.8.13
 Provides-Extra: postgres
 Requires-Dist: psycopg2-binary; extra == "postgres"
 Requires-Dist: pgvector; extra == "postgres"
@@ -197,7 +197,7 @@ Dynamic: license-file
 
 **+16pp better recall than Mem0 on LoCoMo. 100% stale memory precision. Biologically-inspired memory decay for AI agents.**
 
-Persistent memory for Claude that works like human memory — important things stick, forgotten things fade, outdated facts get demoted automatically.
+Persistent memory for Claude and any MCP-compatible AI — works like human memory. Important things stick, forgotten things fade, outdated facts get pruned automatically.
 
 > Early stage — feedback and ideas welcome.
 
@@ -244,7 +244,7 @@ Importance additionally modulates the decay rate within each category. Memories
 
 ## Setup
 
-**Zero infrastructure required** — uses SQLite out of the box. Two commands and you're done.
+**Zero infrastructure required** — uses DuckDB out of the box. Two commands and you're done.
 
 ### 1. Install
 
@@ -252,11 +252,21 @@ Importance additionally modulates the decay rate within each category. Memories
 pip install yourmemory
 ```
 
-All dependencies are installed automatically. No clone, no separate download steps needed.
+All dependencies installed automatically. No clone, no Docker, no database setup.
 
-### 2. Wire into your AI client
+### 2. Get your config
 
-The database is created automatically at `~/.yourmemory/memories.db` on first use. No `.env` file needed.
+Run this once to get your exact config:
+
+```bash
+yourmemory-path
+```
+
+It prints your full executable path and a ready-to-paste config for any MCP client. Copy it.
+
+### 3. Wire into your AI client
+
+The database is created automatically at `~/.yourmemory/memories.duckdb` on first use.
 
 #### Claude Code
 
@@ -276,26 +286,45 @@ Reload Claude Code (`Cmd+Shift+P` → `Developer: Reload Window`).
 
 #### Cline (VS Code)
 
+VS Code doesn't inherit your shell PATH, so use the **full path** from `yourmemory-path`.
+
 In Cline → **MCP Servers** → **Edit MCP Settings**:
 
 ```json
 {
   "mcpServers": {
     "yourmemory": {
-      "command": "yourmemory",
+      "command": "/full/path/to/yourmemory",
       "args": [],
       "env": {
-        "DATABASE_URL": "",
-        "YOURMEMORY_USER": "your_name"
+        "YOURMEMORY_USER": "your_name",
+        "DATABASE_URL": ""
       }
     }
   }
 }
 ```
 
-Replace `your_name` with your name. Save — Cline will connect automatically.
+Run `yourmemory-path` in terminal — it prints the exact config to paste.
+
+#### Cursor
 
-> **Note:** If Cline can't find the `yourmemory` command, run `which yourmemory` in terminal to get the full path and use that instead.
+Add to `~/.cursor/mcp.json`:
+
+```json
+{
+  "mcpServers": {
+    "yourmemory": {
+      "command": "/full/path/to/yourmemory",
+      "args": [],
+      "env": {
+        "YOURMEMORY_USER": "your_name",
+        "DATABASE_URL": ""
+      }
+    }
+  }
+}
+```
 
 #### Claude Desktop
 
@@ -315,9 +344,9 @@ Restart Claude Desktop.
 
 #### Any MCP-compatible client
 
-YourMemory is a standard stdio MCP server. The command is simply `yourmemory`. Add it to any client that supports MCP servers using the same pattern above.
+YourMemory is a standard stdio MCP server. Works with Claude Code, Claude Desktop, Cline, Cursor, Windsurf, Continue, and Zed. Use the full path from `yourmemory-path` if the client doesn't inherit shell PATH.
 
-### 3. Add memory instructions to your project
+### 4. Add memory instructions to your project
 
 Copy `sample_CLAUDE.md` into your project root as `CLAUDE.md` and replace:
 - `YOUR_NAME` — your name (e.g. `Alice`)
@@ -329,13 +358,19 @@ Claude will now follow the recall → store → update workflow automatically on
 
 ### PostgreSQL (optional — for teams or large datasets)
 
-If you have PostgreSQL + pgvector, create a `.env` file:
+Install with Postgres support:
+
+```bash
+pip install yourmemory[postgres]
+```
+
+Then create a `.env` file:
 
 ```bash
 DATABASE_URL=postgresql://YOUR_USER@localhost:5432/yourmemory
 ```
 
-The backend is selected automatically — `postgresql://` in `DATABASE_URL` → Postgres + pgvector, anything else → SQLite.
+The backend is selected automatically — `postgresql://` in `DATABASE_URL` → Postgres + pgvector, anything else → DuckDB.
 
 **macOS**
 ```bash
@@ -349,8 +384,6 @@ sudo apt install postgresql postgresql-contrib postgresql-16-pgvector
 createdb yourmemory
 ```
 
-> **One-liner setup script** (macOS/Linux): `bash scripts/setup_db.sh` handles install + DB creation automatically.
-
 ---
 
 ## MCP Tools
@@ -402,47 +435,20 @@ Runs automatically every 24 hours on startup — no cron needed. Memories below
 
 ---
 
-## REST API
-
-```bash
-# Store
-curl -X POST http://localhost:8000/memories \
-  -H "Content-Type: application/json" \
-  -d '{"userId":"u1","content":"Prefers dark mode","importance":0.8}'
-
-# Retrieve
-curl -X POST http://localhost:8000/retrieve \
-  -H "Content-Type: application/json" \
-  -d '{"userId":"u1","query":"UI preferences"}'
-
-# List all
-curl "http://localhost:8000/memories?userId=u1"
-
-# Update
-curl -X PUT http://localhost:8000/memories/42 \
-  -H "Content-Type: application/json" \
-  -d '{"content":"Prefers dark mode in all apps","importance":0.85}'
-
-# Delete
-curl -X DELETE http://localhost:8000/memories/42
-```
-
----
-
 ## Stack
 
-- **PostgreSQL + pgvector** — vector similarity search
+- **DuckDB** — default backend, zero setup, native vector similarity (same quality as pgvector)
 - **sentence-transformers** — local embeddings (`all-mpnet-base-v2`, 768 dims, no external service needed)
-- **FastAPI** — REST server
 - **APScheduler** — automatic 24h decay job
 - **MCP** — Claude integration via Model Context Protocol
+- **PostgreSQL + pgvector** — optional, for teams / large datasets
 
 ---
 
 ## Architecture
 
 ```
-Claude Code
+Claude / Cline / Cursor / Any MCP client
     │
     ├── recall_memory(query)
     │       └── embed → cosine similarity → score = sim × strength → top-k
@@ -455,12 +461,12 @@ Claude Code
     └── update_memory(id, new_content)
             └── embed(new_content) → UPDATE memories
 
-PostgreSQL (pgvector)
-    └── memories
-        ├── embedding vector(768)
-        ├── importance float
-        ├── recall_count int
-        └── last_accessed_at
+DuckDB (default)                    PostgreSQL + pgvector (optional)
+    └── memories.duckdb                 └── memories table
+        ├── embedding FLOAT[768]            ├── embedding vector(768)
+        ├── importance FLOAT               ├── importance float
+        ├── recall_count INTEGER           ├── recall_count int
+        └── last_accessed_at               └── last_accessed_at
 ```
 
 ---

{yourmemory-1.2.0 → yourmemory-1.2.1}/README.md

@@ -2,7 +2,7 @@
 
 **+16pp better recall than Mem0 on LoCoMo. 100% stale memory precision. Biologically-inspired memory decay for AI agents.**
 
-Persistent memory for Claude that works like human memory — important things stick, forgotten things fade, outdated facts get demoted automatically.
+Persistent memory for Claude and any MCP-compatible AI — works like human memory. Important things stick, forgotten things fade, outdated facts get pruned automatically.
 
 > Early stage — feedback and ideas welcome.
 
@@ -49,7 +49,7 @@ Importance additionally modulates the decay rate within each category. Memories
 
 ## Setup
 
-**Zero infrastructure required** — uses SQLite out of the box. Two commands and you're done.
+**Zero infrastructure required** — uses DuckDB out of the box. Two commands and you're done.
 
 ### 1. Install
 
@@ -57,11 +57,21 @@ Importance additionally modulates the decay rate within each category. Memories
 pip install yourmemory
 ```
 
-All dependencies are installed automatically. No clone, no separate download steps needed.
+All dependencies installed automatically. No clone, no Docker, no database setup.
 
-### 2. Wire into your AI client
+### 2. Get your config
 
-The database is created automatically at `~/.yourmemory/memories.db` on first use. No `.env` file needed.
+Run this once to get your exact config:
+
+```bash
+yourmemory-path
+```
+
+It prints your full executable path and a ready-to-paste config for any MCP client. Copy it.
+
+### 3. Wire into your AI client
+
+The database is created automatically at `~/.yourmemory/memories.duckdb` on first use.
 
 #### Claude Code
 
@@ -81,26 +91,45 @@ Reload Claude Code (`Cmd+Shift+P` → `Developer: Reload Window`).
 
 #### Cline (VS Code)
 
+VS Code doesn't inherit your shell PATH, so use the **full path** from `yourmemory-path`.
+
 In Cline → **MCP Servers** → **Edit MCP Settings**:
 
 ```json
 {
   "mcpServers": {
     "yourmemory": {
-      "command": "yourmemory",
+      "command": "/full/path/to/yourmemory",
       "args": [],
       "env": {
-        "DATABASE_URL": "",
-        "YOURMEMORY_USER": "your_name"
+        "YOURMEMORY_USER": "your_name",
+        "DATABASE_URL": ""
       }
     }
   }
 }
 ```
 
-Replace `your_name` with your name. Save — Cline will connect automatically.
+Run `yourmemory-path` in terminal — it prints the exact config to paste.
+
+#### Cursor
 
-> **Note:** If Cline can't find the `yourmemory` command, run `which yourmemory` in terminal to get the full path and use that instead.
+Add to `~/.cursor/mcp.json`:
+
+```json
+{
+  "mcpServers": {
+    "yourmemory": {
+      "command": "/full/path/to/yourmemory",
+      "args": [],
+      "env": {
+        "YOURMEMORY_USER": "your_name",
+        "DATABASE_URL": ""
+      }
+    }
+  }
+}
+```
 
 #### Claude Desktop
 
@@ -120,9 +149,9 @@ Restart Claude Desktop.
 
 #### Any MCP-compatible client
 
-YourMemory is a standard stdio MCP server. The command is simply `yourmemory`. Add it to any client that supports MCP servers using the same pattern above.
+YourMemory is a standard stdio MCP server. Works with Claude Code, Claude Desktop, Cline, Cursor, Windsurf, Continue, and Zed. Use the full path from `yourmemory-path` if the client doesn't inherit shell PATH.
 
-### 3. Add memory instructions to your project
+### 4. Add memory instructions to your project
 
 Copy `sample_CLAUDE.md` into your project root as `CLAUDE.md` and replace:
 - `YOUR_NAME` — your name (e.g. `Alice`)
@@ -134,13 +163,19 @@ Claude will now follow the recall → store → update workflow automatically on
 
 ### PostgreSQL (optional — for teams or large datasets)
 
-If you have PostgreSQL + pgvector, create a `.env` file:
+Install with Postgres support:
+
+```bash
+pip install yourmemory[postgres]
+```
+
+Then create a `.env` file:
 
 ```bash
 DATABASE_URL=postgresql://YOUR_USER@localhost:5432/yourmemory
 ```
 
-The backend is selected automatically — `postgresql://` in `DATABASE_URL` → Postgres + pgvector, anything else → SQLite.
+The backend is selected automatically — `postgresql://` in `DATABASE_URL` → Postgres + pgvector, anything else → DuckDB.
 
 **macOS**
 ```bash
@@ -154,8 +189,6 @@ sudo apt install postgresql postgresql-contrib postgresql-16-pgvector
 createdb yourmemory
 ```
 
-> **One-liner setup script** (macOS/Linux): `bash scripts/setup_db.sh` handles install + DB creation automatically.
-
 ---
 
 ## MCP Tools
@@ -207,47 +240,20 @@ Runs automatically every 24 hours on startup — no cron needed. Memories below
 
 ---
 
-## REST API
-
-```bash
-# Store
-curl -X POST http://localhost:8000/memories \
-  -H "Content-Type: application/json" \
-  -d '{"userId":"u1","content":"Prefers dark mode","importance":0.8}'
-
-# Retrieve
-curl -X POST http://localhost:8000/retrieve \
-  -H "Content-Type: application/json" \
-  -d '{"userId":"u1","query":"UI preferences"}'
-
-# List all
-curl "http://localhost:8000/memories?userId=u1"
-
-# Update
-curl -X PUT http://localhost:8000/memories/42 \
-  -H "Content-Type: application/json" \
-  -d '{"content":"Prefers dark mode in all apps","importance":0.85}'
-
-# Delete
-curl -X DELETE http://localhost:8000/memories/42
-```
-
----
-
 ## Stack
 
-- **PostgreSQL + pgvector** — vector similarity search
+- **DuckDB** — default backend, zero setup, native vector similarity (same quality as pgvector)
 - **sentence-transformers** — local embeddings (`all-mpnet-base-v2`, 768 dims, no external service needed)
-- **FastAPI** — REST server
 - **APScheduler** — automatic 24h decay job
 - **MCP** — Claude integration via Model Context Protocol
+- **PostgreSQL + pgvector** — optional, for teams / large datasets
 
 ---
 
 ## Architecture
 
 ```
-Claude Code
+Claude / Cline / Cursor / Any MCP client
     │
     ├── recall_memory(query)
     │       └── embed → cosine similarity → score = sim × strength → top-k
@@ -260,12 +266,12 @@ Claude Code
     └── update_memory(id, new_content)
             └── embed(new_content) → UPDATE memories
 
-PostgreSQL (pgvector)
-    └── memories
-        ├── embedding vector(768)
-        ├── importance float
-        ├── recall_count int
-        └── last_accessed_at
+DuckDB (default)                    PostgreSQL + pgvector (optional)
+    └── memories.duckdb                 └── memories table
+        ├── embedding FLOAT[768]            ├── embedding vector(768)
+        ├── importance FLOAT               ├── importance float
+        ├── recall_count INTEGER           ├── recall_count int
+        └── last_accessed_at               └── last_accessed_at
 ```
 
 ---

{yourmemory-1.2.0 → yourmemory-1.2.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "yourmemory"
-version = "1.2.0"
+version = "1.2.1"
 description = "Persistent memory for Claude — Ebbinghaus forgetting curve, semantic deduplication, MCP-native"
 readme = "README.md"
 requires-python = ">=3.11"
@@ -28,7 +28,7 @@ dependencies = [
     "python-dateutil",
     "duckdb>=0.10.0",
     "apscheduler",
-    "spacy>=3.7,<4.0",
+    "spacy>=3.8.13,<4.0",
 ]
 
 [project.optional-dependencies]

{yourmemory-1.2.0 → yourmemory-1.2.1}/src/services/extract.py

@@ -29,9 +29,7 @@ def categorize(text: str) -> str:
     """
     Use spaCy dependency parse to classify:
       fact       — declarative sentence with an explicit subject
-                   e.g. "Novak is world number 1", "I love Python", "My name is Sachit"
       assumption — imperative sentence with no subject (command/instruction)
-                   e.g. "Use python instead of JS", "Please convert this image"
     """
     doc = _nlp(text)
     has_subject = any(tok.dep_ in ("nsubj", "nsubjpass") for tok in doc)

yourmemory-1.2.1/src/services/extract_fallback.py

@@ -0,0 +1,34 @@
+import re
+
+_QUESTION_WORDS = {"what", "who", "where", "when", "why", "how", "which", "whose", "whom"}
+
+
+def is_question(text: str) -> bool:
+    """Return True if the text is a question — questions are not stored as memories."""
+    stripped = text.strip()
+    if stripped.endswith("?"):
+        return True
+    first_word = re.split(r"\s+", stripped.lower())[0]
+    return first_word in _QUESTION_WORDS
+
+
+def categorize(text: str) -> str:
+    """
+    Fallback categorization without spaCy:
+    Simple heuristics to classify as fact vs assumption
+    """
+    text_lower = text.lower().strip()
+    
+    # Commands/imperatives typically start with verbs or "please"
+    imperative_patterns = [
+        r'^(please|use|try|do|don\'t|make|create|add|remove|delete|update)',
+        r'^(convert|transform|change|modify|fix|help|show|tell)',
+        r'^(install|run|execute|start|stop|restart|configure)'
+    ]
+    
+    for pattern in imperative_patterns:
+        if re.match(pattern, text_lower):
+            return "assumption"
+    
+    # Default to fact for declarative statements
+    return "fact"

{yourmemory-1.2.0 → yourmemory-1.2.1/yourmemory.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: yourmemory
-Version: 1.2.0
+Version: 1.2.1
 Summary: Persistent memory for Claude — Ebbinghaus forgetting curve, semantic deduplication, MCP-native
 Author-email: Sachit Misra <mishrasachit1@gmail.com>
 License:                                  Apache License
@@ -181,7 +181,7 @@ Requires-Dist: numpy
 Requires-Dist: python-dateutil
 Requires-Dist: duckdb>=0.10.0
 Requires-Dist: apscheduler
-Requires-Dist: spacy<4.0,>=3.7
+Requires-Dist: spacy<4.0,>=3.8.13
 Provides-Extra: postgres
 Requires-Dist: psycopg2-binary; extra == "postgres"
 Requires-Dist: pgvector; extra == "postgres"
@@ -197,7 +197,7 @@ Dynamic: license-file
 
 **+16pp better recall than Mem0 on LoCoMo. 100% stale memory precision. Biologically-inspired memory decay for AI agents.**
 
-Persistent memory for Claude that works like human memory — important things stick, forgotten things fade, outdated facts get demoted automatically.
+Persistent memory for Claude and any MCP-compatible AI — works like human memory. Important things stick, forgotten things fade, outdated facts get pruned automatically.
 
 > Early stage — feedback and ideas welcome.
 
@@ -244,7 +244,7 @@ Importance additionally modulates the decay rate within each category. Memories
 
 ## Setup
 
-**Zero infrastructure required** — uses SQLite out of the box. Two commands and you're done.
+**Zero infrastructure required** — uses DuckDB out of the box. Two commands and you're done.
 
 ### 1. Install
 
@@ -252,11 +252,21 @@ Importance additionally modulates the decay rate within each category. Memories
 pip install yourmemory
 ```
 
-All dependencies are installed automatically. No clone, no separate download steps needed.
+All dependencies installed automatically. No clone, no Docker, no database setup.
 
-### 2. Wire into your AI client
+### 2. Get your config
 
-The database is created automatically at `~/.yourmemory/memories.db` on first use. No `.env` file needed.
+Run this once to get your exact config:
+
+```bash
+yourmemory-path
+```
+
+It prints your full executable path and a ready-to-paste config for any MCP client. Copy it.
+
+### 3. Wire into your AI client
+
+The database is created automatically at `~/.yourmemory/memories.duckdb` on first use.
 
 #### Claude Code
 
@@ -276,26 +286,45 @@ Reload Claude Code (`Cmd+Shift+P` → `Developer: Reload Window`).
 
 #### Cline (VS Code)
 
+VS Code doesn't inherit your shell PATH, so use the **full path** from `yourmemory-path`.
+
 In Cline → **MCP Servers** → **Edit MCP Settings**:
 
 ```json
 {
   "mcpServers": {
     "yourmemory": {
-      "command": "yourmemory",
+      "command": "/full/path/to/yourmemory",
       "args": [],
       "env": {
-        "DATABASE_URL": "",
-        "YOURMEMORY_USER": "your_name"
+        "YOURMEMORY_USER": "your_name",
+        "DATABASE_URL": ""
       }
     }
   }
 }
 ```
 
-Replace `your_name` with your name. Save — Cline will connect automatically.
+Run `yourmemory-path` in terminal — it prints the exact config to paste.
+
+#### Cursor
 
-> **Note:** If Cline can't find the `yourmemory` command, run `which yourmemory` in terminal to get the full path and use that instead.
+Add to `~/.cursor/mcp.json`:
+
+```json
+{
+  "mcpServers": {
+    "yourmemory": {
+      "command": "/full/path/to/yourmemory",
+      "args": [],
+      "env": {
+        "YOURMEMORY_USER": "your_name",
+        "DATABASE_URL": ""
+      }
+    }
+  }
+}
+```
 
 #### Claude Desktop
 
@@ -315,9 +344,9 @@ Restart Claude Desktop.
 
 #### Any MCP-compatible client
 
-YourMemory is a standard stdio MCP server. The command is simply `yourmemory`. Add it to any client that supports MCP servers using the same pattern above.
+YourMemory is a standard stdio MCP server. Works with Claude Code, Claude Desktop, Cline, Cursor, Windsurf, Continue, and Zed. Use the full path from `yourmemory-path` if the client doesn't inherit shell PATH.
 
-### 3. Add memory instructions to your project
+### 4. Add memory instructions to your project
 
 Copy `sample_CLAUDE.md` into your project root as `CLAUDE.md` and replace:
 - `YOUR_NAME` — your name (e.g. `Alice`)
@@ -329,13 +358,19 @@ Claude will now follow the recall → store → update workflow automatically on
 
 ### PostgreSQL (optional — for teams or large datasets)
 
-If you have PostgreSQL + pgvector, create a `.env` file:
+Install with Postgres support:
+
+```bash
+pip install yourmemory[postgres]
+```
+
+Then create a `.env` file:
 
 ```bash
 DATABASE_URL=postgresql://YOUR_USER@localhost:5432/yourmemory
 ```
 
-The backend is selected automatically — `postgresql://` in `DATABASE_URL` → Postgres + pgvector, anything else → SQLite.
+The backend is selected automatically — `postgresql://` in `DATABASE_URL` → Postgres + pgvector, anything else → DuckDB.
 
 **macOS**
 ```bash
@@ -349,8 +384,6 @@ sudo apt install postgresql postgresql-contrib postgresql-16-pgvector
 createdb yourmemory
 ```
 
-> **One-liner setup script** (macOS/Linux): `bash scripts/setup_db.sh` handles install + DB creation automatically.
-
 ---
 
 ## MCP Tools
@@ -402,47 +435,20 @@ Runs automatically every 24 hours on startup — no cron needed. Memories below
 
 ---
 
-## REST API
-
-```bash
-# Store
-curl -X POST http://localhost:8000/memories \
-  -H "Content-Type: application/json" \
-  -d '{"userId":"u1","content":"Prefers dark mode","importance":0.8}'
-
-# Retrieve
-curl -X POST http://localhost:8000/retrieve \
-  -H "Content-Type: application/json" \
-  -d '{"userId":"u1","query":"UI preferences"}'
-
-# List all
-curl "http://localhost:8000/memories?userId=u1"
-
-# Update
-curl -X PUT http://localhost:8000/memories/42 \
-  -H "Content-Type: application/json" \
-  -d '{"content":"Prefers dark mode in all apps","importance":0.85}'
-
-# Delete
-curl -X DELETE http://localhost:8000/memories/42
-```
-
----
-
 ## Stack
 
-- **PostgreSQL + pgvector** — vector similarity search
+- **DuckDB** — default backend, zero setup, native vector similarity (same quality as pgvector)
 - **sentence-transformers** — local embeddings (`all-mpnet-base-v2`, 768 dims, no external service needed)
-- **FastAPI** — REST server
 - **APScheduler** — automatic 24h decay job
 - **MCP** — Claude integration via Model Context Protocol
+- **PostgreSQL + pgvector** — optional, for teams / large datasets
 
 ---
 
 ## Architecture
 
 ```
-Claude Code
+Claude / Cline / Cursor / Any MCP client
     │
     ├── recall_memory(query)
     │       └── embed → cosine similarity → score = sim × strength → top-k
@@ -455,12 +461,12 @@ Claude Code
     └── update_memory(id, new_content)
             └── embed(new_content) → UPDATE memories
 
-PostgreSQL (pgvector)
-    └── memories
-        ├── embedding vector(768)
-        ├── importance float
-        ├── recall_count int
-        └── last_accessed_at
+DuckDB (default)                    PostgreSQL + pgvector (optional)
+    └── memories.duckdb                 └── memories table
+        ├── embedding FLOAT[768]            ├── embedding vector(768)
+        ├── importance FLOAT               ├── importance float
+        ├── recall_count INTEGER           ├── recall_count int
+        └── last_accessed_at               └── last_accessed_at
 ```
 
 ---

{yourmemory-1.2.0 → yourmemory-1.2.1}/yourmemory.egg-info/SOURCES.txt

@@ -20,6 +20,7 @@ src/services/api_keys.py
 src/services/decay.py
 src/services/embed.py
 src/services/extract.py
+src/services/extract_fallback.py
 src/services/resolve.py
 src/services/retrieve.py
 yourmemory.egg-info/PKG-INFO

{yourmemory-1.2.0 → yourmemory-1.2.1}/yourmemory.egg-info/requires.txt

@@ -5,7 +5,7 @@ numpy
 python-dateutil
 duckdb>=0.10.0
 apscheduler
-spacy<4.0,>=3.7
+spacy<4.0,>=3.8.13
 
 [all]
 yourmemory[postgres,sse]