feat: added a basic tool to search past conversation logs (#109)

Author: ErikBjare

gptme/tools/__init__.py

@@ -14,6 +14,7 @@
 from .python import register_function
 from .read import tool as tool_read
 from .save import execute_save, tool_append, tool_save
+from .search_chats import tool as search_chats_tool
 from .shell import execute_shell
 from .shell import tool as shell_tool
 from .subagent import tool as subagent_tool
@@ -44,6 +45,7 @@
     tmux_tool,
     browser_tool,
     gh_tool,
+    search_chats_tool,
     # python tool is loaded last to ensure all functions are registered
     get_python_tool,
 ]

gptme/tools/search_chats.py

@@ -0,0 +1,87 @@
+import logging
+from pathlib import Path
+
+from .base import ToolSpec
+
+logger = logging.getLogger(__name__)
+
+
+def search_chats(query: str, max_results: int = 5) -> None:
+    """
+    Search past conversation logs for the given query and print a summary of the results.
+
+    Args:
+        query (str): The search query.
+        max_results (int): Maximum number of conversations to display.
+    """
+    # noreorder
+    from ..logmanager import LogManager, get_conversations  # fmt: skip
+
+    conversations = list(get_conversations())
+    results = []
+
+    for conv in conversations:
+        log_path = Path(conv["path"])
+        log_manager = LogManager.load(log_path)
+
+        matching_messages = []
+        for msg in log_manager.log:
+            if query.lower() in msg.content.lower():
+                matching_messages.append(msg)
+
+        if matching_messages:
+            results.append(
+                {
+                    "conversation": conv["name"],
+                    "messages": matching_messages,
+                }
+            )
+
+    # Sort results by the number of matching messages, in descending order
+    results.sort(key=lambda x: len(x["messages"]), reverse=True)
+    results = results[:max_results]
+
+    if not results:
+        print(f"No results found for query: '{query}'")
+        return
+
+    print(f"Search results for query: '{query}'")
+    print(f"Found matches in {len(results)} conversation(s):")
+
+    for i, result in enumerate(results, 1):
+        print(f"\n{i}. Conversation: {result['conversation']}")
+        print(f"   Number of matching messages: {len(result['messages'])}")
+        print("   Sample matches:")
+        # Show up to 3 sample messages
+        for j, msg in enumerate(result["messages"][:3], 1):
+            content = (
+                msg.content[:100] + "..." if len(msg.content) > 100 else msg.content
+            )
+            print(f"     {j}. {msg.role.capitalize()}: {content}")
+        if len(result["messages"]) > 3:
+            print(
+                f"     ... and {len(result['messages']) - 3} more matching message(s)"
+            )
+
+
+instructions = """
+To search past conversation logs, you can use the `search_chats` function in Python.
+This function allows you to find relevant information from previous conversations.
+"""
+
+examples = """
+### Search for a specific topic in past conversations
+User: Can you find any mentions of "python" in our past conversations?
+Assistant: Certainly! I'll search our past conversations for mentions of "python" using the search_chats function.
+```python
+search_chats("python")
+```
+"""
+
+tool = ToolSpec(
+    name="search_chats",
+    desc="Search past conversation logs",
+    instructions=instructions,
+    examples=examples,
+    functions=[search_chats],
+)