cognify-code 0.2.5__py3-none-any.whl → 0.2.6__py3-none-any.whl

ai_code_assistant/agent/code_agent.py

@@ -808,339 +808,6 @@ Provide a helpful, concise response.
         response = AgentResponse(message=full_response, intent=intent)
         yield ("", response)
 
-
-
-    def process_stream(self, message: str, use_llm_classification: bool = True) -> Iterator[Tuple[str, Optional[AgentResponse]]]:
-        """
-        Process a user message with streaming output.
-        
-        Yields tuples of (chunk, final_response).
-        During streaming, final_response is None.
-        The last yield will have the complete AgentResponse.
-        """
-        # Classify intent (non-streaming, it's fast)
-        if use_llm_classification:
-            intent = self.intent_classifier.classify_with_llm(message)
-        else:
-            intent = self.intent_classifier.classify(message)
-        
-        # Route to appropriate streaming handler
-        streaming_handlers = {
-            IntentType.CODE_GENERATE: self._handle_generate_stream,
-            IntentType.CODE_EDIT: self._handle_edit_stream,
-            IntentType.CODE_REVIEW: self._handle_review_stream,
-            IntentType.CODE_EXPLAIN: self._handle_explain_stream,
-            IntentType.CODE_REFACTOR: self._handle_refactor_stream,
-            IntentType.GENERAL_CHAT: self._handle_general_chat_stream,
-        }
-        
-        handler = streaming_handlers.get(intent.type)
-        
-        if handler:
-            yield from handler(message, intent)
-        else:
-            # Fall back to non-streaming for other intents
-            response = self.process(message, use_llm_classification)
-            yield (response.message, response)
-    
-    def _handle_explain_stream(self, message: str, intent: Intent) -> Iterator[Tuple[str, Optional[AgentResponse]]]:
-        """Handle code explanation with streaming."""
-        if not intent.file_paths:
-            response = AgentResponse(
-                message="Please specify which file or code you want me to explain.",
-                intent=intent,
-            )
-            yield (response.message, response)
-            return
-        
-        file_path = intent.file_paths[0]
-        content = self.file_manager.read_file(file_path)
-        
-        if not content:
-            response = AgentResponse(
-                message=f"Cannot find file: {file_path}",
-                intent=intent,
-            )
-            yield (response.message, response)
-            return
-        
-        prompt = f"""Explain the following code in a clear, educational way.
-
-## Code ({file_path})
-```
-{content[:5000]}
-```
-
-## Instructions
-1. Start with a high-level overview
-2. Explain the main components/functions
-3. Describe the flow of execution
-4. Note any important patterns or techniques used
-5. Keep the explanation concise but thorough
-"""
-        
-        # Stream the explanation
-        full_response = f"📖 **Explanation of {file_path}**\n\n"
-        yield (f"📖 **Explanation of {file_path}**\n\n", None)
-        
-        for chunk in self.llm.stream(prompt):
-            full_response += chunk
-            yield (chunk, None)
-        
-        # Final response
-        response = AgentResponse(
-            message=full_response,
-            intent=intent,
-        )
-        yield ("", response)
-    
-    def _handle_review_stream(self, message: str, intent: Intent) -> Iterator[Tuple[str, Optional[AgentResponse]]]:
-        """Handle code review with streaming."""
-        if not intent.file_paths:
-            context = self.file_manager.get_project_context()
-            py_files = [f.relative_path for f in context.files if f.extension == ".py"][:5]
-            
-            if py_files:
-                msg = f"Which file would you like me to review? Found these Python files:\n" + \
-                      "\n".join(f"  • {f}" for f in py_files)
-            else:
-                msg = "Please specify which file you want me to review."
-            
-            response = AgentResponse(message=msg, intent=intent)
-            yield (msg, response)
-            return
-        
-        file_path = intent.file_paths[0]
-        content = self.file_manager.read_file(file_path)
-        
-        if not content:
-            response = AgentResponse(
-                message=f"Cannot find file: {file_path}",
-                intent=intent,
-            )
-            yield (response.message, response)
-            return
-        
-        yield (f"🔍 **Reviewing {file_path}...**\n\n", None)
-        
-        # Use streaming for the review
-        prompt = f"""Review the following code for issues, bugs, and improvements.
-
-## Code ({file_path})
-```
-{content[:5000]}
-```
-
-## Review Format
-Provide a structured review with:
-1. **Summary** - Brief overview
-2. **Issues** - List any bugs, security issues, or problems
-3. **Suggestions** - Improvements and best practices
-4. **Score** - Rate the code quality (1-10)
-
-Be specific and actionable.
-"""
-        
-        full_response = f"🔍 **Reviewing {file_path}...**\n\n"
-        
-        for chunk in self.llm.stream(prompt):
-            full_response += chunk
-            yield (chunk, None)
-        
-        response = AgentResponse(
-            message=full_response,
-            intent=intent,
-        )
-        yield ("", response)
-    
-    def _handle_generate_stream(self, message: str, intent: Intent) -> Iterator[Tuple[str, Optional[AgentResponse]]]:
-        """Handle code generation with streaming."""
-        yield ("🔨 **Generating code...**\n\n", None)
-        
-        request = CodeGenerationRequest(
-            description=message,
-            language=intent.language,
-            file_path=intent.file_paths[0] if intent.file_paths else None,
-        )
-        
-        # Generate code (this part streams)
-        full_code = ""
-        prompt = self.code_generator._build_prompt(request)
-        
-        for chunk in self.llm.stream(prompt):
-            full_code += chunk
-            yield (chunk, None)
-        
-        # Extract and create changeset
-        code = self._extract_code(full_code)
-        file_path = request.file_path or f"generated.{request.language or 'py'}"
-        
-        generated = GeneratedCode(
-            code=code,
-            language=request.language or "python",
-            file_path=file_path,
-            description=request.description,
-        )
-        
-        changeset = ChangeSet(description=f"Generate: {message[:50]}...")
-        diff = self.diff_engine.create_file_diff(generated.file_path, generated.code)
-        changeset.diffs.append(diff)
-        
-        self._pending_changeset = changeset
-        
-        response = AgentResponse(
-            message=f"\n\n✅ Code generated for {file_path}",
-            intent=intent,
-            generated_code=generated,
-            changeset=changeset,
-            requires_confirmation=True,
-        )
-        yield ("\n\n✅ Code generated. Apply changes? (yes/no)", response)
-    
-    def _handle_edit_stream(self, message: str, intent: Intent) -> Iterator[Tuple[str, Optional[AgentResponse]]]:
-        """Handle code editing with streaming."""
-        if not intent.file_paths:
-            response = AgentResponse(
-                message="Please specify which file you want to edit.",
-                intent=intent,
-            )
-            yield (response.message, response)
-            return
-        
-        file_path = intent.file_paths[0]
-        original = self.file_manager.read_file(file_path)
-        
-        if not original:
-            response = AgentResponse(
-                message=f"Cannot find file: {file_path}",
-                intent=intent,
-            )
-            yield (response.message, response)
-            return
-        
-        yield (f"✏️ **Editing {file_path}...**\n\n", None)
-        
-        prompt = f"""Edit the following code according to the user's request.
-
-## Original Code ({file_path})
-```
-{original[:5000]}
-```
-
-## User Request
-{message}
-
-## Instructions
-Return the COMPLETE modified file.
-
-```
-"""
-        
-        full_response = ""
-        for chunk in self.llm.stream(prompt):
-            full_response += chunk
-            yield (chunk, None)
-        
-        new_code = self._extract_code(full_response)
-        
-        changeset = ChangeSet(description=f"Edit: {message[:50]}...")
-        diff = self.diff_engine.create_diff(original, new_code, file_path)
-        changeset.diffs.append(diff)
-        
-        self._pending_changeset = changeset
-        
-        response = AgentResponse(
-            message=f"\n\n✅ Edit complete for {file_path}",
-            intent=intent,
-            changeset=changeset,
-            requires_confirmation=True,
-        )
-        yield ("\n\n✅ Edit complete. Apply changes? (yes/no)", response)
-    
-    def _handle_refactor_stream(self, message: str, intent: Intent) -> Iterator[Tuple[str, Optional[AgentResponse]]]:
-        """Handle code refactoring with streaming."""
-        if not intent.file_paths:
-            response = AgentResponse(
-                message="Please specify which file you want to refactor.",
-                intent=intent,
-            )
-            yield (response.message, response)
-            return
-        
-        file_path = intent.file_paths[0]
-        original = self.file_manager.read_file(file_path)
-        
-        if not original:
-            response = AgentResponse(
-                message=f"Cannot find file: {file_path}",
-                intent=intent,
-            )
-            yield (response.message, response)
-            return
-        
-        yield (f"🔄 **Refactoring {file_path}...**\n\n", None)
-        
-        prompt = f"""Refactor the following code to improve its quality.
-
-## Original Code ({file_path})
-```
-{original[:5000]}
-```
-
-## User Request
-{message}
-
-Return the COMPLETE refactored file.
-
-```
-"""
-        
-        full_response = ""
-        for chunk in self.llm.stream(prompt):
-            full_response += chunk
-            yield (chunk, None)
-        
-        new_code = self._extract_code(full_response)
-        
-        changeset = ChangeSet(description=f"Refactor: {file_path}")
-        diff = self.diff_engine.create_diff(original, new_code, file_path)
-        changeset.diffs.append(diff)
-        
-        self._pending_changeset = changeset
-        
-        response = AgentResponse(
-            message=f"\n\n✅ Refactoring complete for {file_path}",
-            intent=intent,
-            changeset=changeset,
-            requires_confirmation=True,
-        )
-        yield ("\n\n✅ Refactoring complete. Apply changes? (yes/no)", response)
-    
-    def _handle_general_chat_stream(self, message: str, intent: Intent) -> Iterator[Tuple[str, Optional[AgentResponse]]]:
-        """Handle general chat with streaming."""
-        context = self.file_manager.get_project_context()
-        
-        prompt = f"""You are a helpful coding assistant. Answer the user's question.
-
-Project context:
-- Root: {context.root_path.name}
-- Languages: {', '.join(context.languages)}
-- Files: {context.total_code_files} code files
-
-User: {message}
-
-Provide a helpful, concise response.
-"""
-        
-        full_response = ""
-        for chunk in self.llm.stream(prompt):
-            full_response += chunk
-            yield (chunk, None)
-        
-        response = AgentResponse(message=full_response, intent=intent)
-        yield ("", response)
-
-
     def _extract_code(self, response: str) -> str:
         """Extract code from LLM response."""
         import re

ai_code_assistant/cli.py

@@ -66,6 +66,59 @@ def get_components(config_path: Optional[Path] = None):
     return config, llm
 
 
+def validate_llm_connection(llm: LLMManager, exit_on_failure: bool = True) -> bool:
+    """
+    Validate LLM connection before operations.
+
+    Args:
+        llm: The LLM manager instance
+        exit_on_failure: If True, exit with error code on failure
+
+    Returns:
+        True if connection is valid, False otherwise
+    """
+    try:
+        info = llm.get_model_info()
+        provider = info.get("provider", "ollama")
+
+        # Quick validation - don't do full connection check for cloud providers with API keys
+        if provider != "ollama" and llm.config.llm.api_key:
+            return True
+
+        # For Ollama, check connection
+        if provider == "ollama":
+            # Try a lightweight check first
+            import socket
+            base_url = info.get("base_url", "http://localhost:11434")
+            host = base_url.replace("http://", "").replace("https://", "").split(":")[0]
+            port = int(base_url.split(":")[-1].split("/")[0]) if ":" in base_url else 11434
+
+            sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+            sock.settimeout(2)
+            result = sock.connect_ex((host, port))
+            sock.close()
+
+            if result != 0:
+                console.print(f"\n[red]Error: Cannot connect to Ollama at {base_url}[/red]")
+                console.print("\n[yellow]Quick fix:[/yellow]")
+                console.print("  1. Make sure Ollama is installed: [cyan]https://ollama.ai[/cyan]")
+                console.print(f"  2. Pull the model: [cyan]ollama pull {info.get('model', 'deepseek-coder:6.7b')}[/cyan]")
+                console.print("  3. Start Ollama: [cyan]ollama serve[/cyan]")
+                console.print("\n[dim]Or use a cloud provider: cognify status --help[/dim]")
+
+                if exit_on_failure:
+                    sys.exit(1)
+                return False
+
+        return True
+
+    except Exception as e:
+        console.print(f"\n[red]Error validating LLM connection: {e}[/red]")
+        if exit_on_failure:
+            sys.exit(1)
+        return False
+
+
 @click.group(invoke_without_command=True)
 @click.version_option(version=__version__, prog_name="cognify")
 @click.option("--config", "-c", type=click.Path(exists=True, path_type=Path), help="Config file path")
@@ -104,6 +157,7 @@ def review(ctx, files: Tuple[Path, ...], review_type: str, output_format: str,
         sys.exit(1)
 
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     analyzer = CodeAnalyzer(config, llm)
     file_handler = FileHandler(config)
     formatter = get_formatter(output_format, config.output.use_colors)
@@ -178,6 +232,7 @@ def generate(ctx, description: str, mode: str, language: str, name: Optional[str
     from rich.panel import Panel
     
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     generator = CodeGenerator(config, llm)
     formatter = get_formatter(output_format, config.output.use_colors)
 
@@ -267,6 +322,7 @@ def generate(ctx, description: str, mode: str, language: str, name: Optional[str
 def chat(ctx, context: Tuple[Path, ...], stream: bool):
     """Start an interactive chat session about code."""
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     session = ChatSession(config, llm)
 
     # Load context files
@@ -558,6 +614,7 @@ def edit(ctx, file: Path, instruction: str, mode: str, preview: bool,
         ai-assist edit config.py "Update the timeout value" -s 10 -e 20
     """
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     editor = FileEditor(config, llm)
 
     # Determine edit mode
@@ -647,6 +704,7 @@ def refactor(ctx, instruction: str, files: Tuple[Path, ...], pattern: Optional[s
     from ai_code_assistant.utils import FileHandler
 
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     editor = MultiFileEditor(config, llm)
     file_handler = FileHandler(config)
 
@@ -788,6 +846,7 @@ def rename(ctx, old_name: str, new_name: str, symbol_type: str, files: Tuple[Pat
     from ai_code_assistant.utils import FileHandler
 
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     editor = MultiFileEditor(config, llm)
     file_handler = FileHandler(config)
 
@@ -1182,6 +1241,7 @@ def git_commit(ctx, message: Optional[str], stage_all: bool, push_after: bool, n
     from ai_code_assistant.git import GitManager, CommitMessageGenerator
 
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
 
     try:
         git_mgr = GitManager()
@@ -1315,6 +1375,7 @@ def git_sync(ctx, message: Optional[str], no_confirm: bool):
     from ai_code_assistant.git import GitManager, CommitMessageGenerator
 
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
 
     try:
         git_mgr = GitManager()
@@ -1451,9 +1512,10 @@ def agent(ctx, path: Path):
     from ai_code_assistant.chat import AgentChatSession
     from rich.markdown import Markdown
     from rich.prompt import Prompt
-    
+
     config, llm = get_components(ctx.obj.get("config_path"))
-    
+    validate_llm_connection(llm)
+
     # Initialize agent session
     session = AgentChatSession(config, llm, path.resolve())
     
@@ -1591,10 +1653,11 @@ def agent_review(ctx, file: Path, path: Path, stream: bool):
         ai-assist agent-review main.py --no-stream
     """
     from ai_code_assistant.agent import CodeAgent
-    
+
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     agent = CodeAgent(llm, path.resolve())
-    
+
     console.print(f"\n[bold]Reviewing {file}...[/bold]\n")
     
     if stream:
@@ -1628,10 +1691,11 @@ def agent_generate(ctx, description: str, file: Optional[Path], language: Option
         ai-assist agent-generate "hello world" --no-stream
     """
     from ai_code_assistant.agent import CodeAgent
-    
+
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     agent = CodeAgent(llm, path.resolve())
-    
+
     # Build the request
     request = description
     if file:
@@ -1681,10 +1745,11 @@ def agent_explain(ctx, file: Path, path: Path, stream: bool):
         ai-assist agent-explain main.py --no-stream
     """
     from ai_code_assistant.agent import CodeAgent
-    
+
     config, llm = get_components(ctx.obj.get("config_path"))
+    validate_llm_connection(llm)
     agent = CodeAgent(llm, path.resolve())
-    
+
     console.print(f"\n[bold]Explaining {file}...[/bold]\n")
     
     if stream:

ai_code_assistant/providers/base.py

@@ -3,10 +3,20 @@
 from abc import ABC, abstractmethod
 from enum import Enum
 from typing import Any, Dict, Iterator, List, Optional
+import logging
 
 from langchain_core.language_models import BaseChatModel
 from langchain_core.messages import BaseMessage, HumanMessage, SystemMessage
 from pydantic import BaseModel
+from tenacity import (
+    retry,
+    stop_after_attempt,
+    wait_exponential,
+    retry_if_exception_type,
+    before_sleep_log,
+)
+
+logger = logging.getLogger(__name__)
 
 
 class ProviderType(str, Enum):
@@ -39,6 +49,21 @@ class ModelInfo(BaseModel):
     is_free: bool = True
 
 
+def _is_retryable_error(exception: Exception) -> bool:
+    """Check if an exception should trigger a retry."""
+    error_msg = str(exception).lower()
+    # Rate limits, temporary server issues, timeout errors
+    retryable_keywords = [
+        "rate limit", "rate_limit", "429",
+        "503", "502", "504",
+        "timeout", "timed out",
+        "temporarily unavailable",
+        "server error", "internal error",
+        "connection", "network",
+    ]
+    return any(keyword in error_msg for keyword in retryable_keywords)
+
+
 class BaseProvider(ABC):
     """Abstract base class for LLM providers."""
 
@@ -52,6 +77,11 @@ class BaseProvider(ABC):
     # Available models for this provider
     available_models: List[ModelInfo] = []
 
+    # Retry settings for cloud providers
+    _retry_attempts: int = 3
+    _retry_min_wait: int = 2
+    _retry_max_wait: int = 30
+
     def __init__(self, config: ProviderConfig):
         """Initialize the provider with configuration."""
         self.config = config
@@ -74,6 +104,10 @@ class BaseProvider(ABC):
         """Validate the provider configuration. Returns (is_valid, error_message)."""
         pass
 
+    def _should_retry(self) -> bool:
+        """Check if this provider should use retry logic (cloud providers only)."""
+        return self.provider_type != ProviderType.OLLAMA
+
     def invoke(self, prompt: str, system_prompt: Optional[str] = None) -> str:
         """Invoke the LLM with a prompt and optional system message."""
         messages: List[BaseMessage] = []
@@ -81,8 +115,33 @@ class BaseProvider(ABC):
             messages.append(SystemMessage(content=system_prompt))
         messages.append(HumanMessage(content=prompt))
 
-        response = self.llm.invoke(messages)
-        return str(response.content)
+        if self._should_retry():
+            return self._invoke_with_retry(messages)
+        else:
+            response = self.llm.invoke(messages)
+            return str(response.content)
+
+    def _invoke_with_retry(self, messages: List[BaseMessage]) -> str:
+        """Invoke with exponential backoff retry for cloud providers."""
+        @retry(
+            stop=stop_after_attempt(self._retry_attempts),
+            wait=wait_exponential(multiplier=1, min=self._retry_min_wait, max=self._retry_max_wait),
+            retry=retry_if_exception_type(Exception),
+            before_sleep=before_sleep_log(logger, logging.WARNING),
+            reraise=True,
+        )
+        def _do_invoke():
+            try:
+                response = self.llm.invoke(messages)
+                return str(response.content)
+            except Exception as e:
+                if _is_retryable_error(e):
+                    logger.warning(f"Retryable error from {self.provider_type.value}: {e}")
+                    raise  # Will be retried
+                else:
+                    raise  # Non-retryable, will not retry
+
+        return _do_invoke()
 
     def stream(self, prompt: str, system_prompt: Optional[str] = None) -> Iterator[str]:
         """Stream LLM response for real-time output."""
@@ -91,8 +150,39 @@ class BaseProvider(ABC):
             messages.append(SystemMessage(content=system_prompt))
         messages.append(HumanMessage(content=prompt))
 
-        for chunk in self.llm.stream(messages):
-            yield str(chunk.content)
+        if self._should_retry():
+            yield from self._stream_with_retry(messages)
+        else:
+            for chunk in self.llm.stream(messages):
+                yield str(chunk.content)
+
+    def _stream_with_retry(self, messages: List[BaseMessage]) -> Iterator[str]:
+        """Stream with retry logic for cloud providers."""
+        last_exception = None
+
+        for attempt in range(self._retry_attempts):
+            try:
+                for chunk in self.llm.stream(messages):
+                    yield str(chunk.content)
+                return  # Success, exit
+            except Exception as e:
+                last_exception = e
+                if _is_retryable_error(e) and attempt < self._retry_attempts - 1:
+                    import time
+                    wait_time = min(
+                        self._retry_max_wait,
+                        self._retry_min_wait * (2 ** attempt)
+                    )
+                    logger.warning(
+                        f"Retryable error from {self.provider_type.value}: {e}. "
+                        f"Retrying in {wait_time}s (attempt {attempt + 1}/{self._retry_attempts})"
+                    )
+                    time.sleep(wait_time)
+                else:
+                    raise
+
+        if last_exception:
+            raise last_exception
 
     def check_connection(self) -> bool:
         """Check if the provider is accessible."""

{cognify_code-0.2.5.dist-info → cognify_code-0.2.6.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cognify-code
-Version: 0.2.5
+Version: 0.2.6
 Summary: Your local AI-powered code assistant. Review, generate, search, and refactor code with an intelligent AI agent—all running locally with complete privacy.
 Author-email: Ashok Kumar <akkssy@users.noreply.github.com>
 Maintainer-email: Ashok Kumar <akkssy@users.noreply.github.com>
@@ -39,6 +39,7 @@ Requires-Dist: rich>=13.0.0
 Requires-Dist: pyyaml>=6.0
 Requires-Dist: pydantic>=2.0.0
 Requires-Dist: pydantic-settings>=2.0.0
+Requires-Dist: tenacity>=8.0.0
 Requires-Dist: chromadb>=0.4.0
 Requires-Dist: sentence-transformers>=2.0.0
 Requires-Dist: watchdog>=4.0.0
@@ -68,11 +69,11 @@ Dynamic: license-file
 
 <p align="center">
   <a href="#features">Features</a> •
+  <a href="#vscode-extension">VSCode Extension</a> •
   <a href="#installation">Installation</a> •
   <a href="#providers">Providers</a> •
   <a href="#usage">Usage</a> •
-  <a href="#documentation">Docs</a> •
-  <a href="#contributing">Contributing</a>
+  <a href="#documentation">Docs</a>
 </p>
 
 <p align="center">
@@ -80,11 +81,12 @@ Dynamic: license-file
   <img src="https://img.shields.io/badge/license-MIT-green.svg" alt="MIT License">
   <img src="https://img.shields.io/badge/tests-144%20passed-brightgreen.svg" alt="Tests">
   <img src="https://img.shields.io/badge/providers-6%20supported-purple.svg" alt="6 Providers">
+  <img src="https://img.shields.io/badge/VSCode-Extension-blue.svg" alt="VSCode Extension">
 </p>
 
 ---
 
-A powerful CLI tool that brings AI-powered code cognition to your terminal. Review code, generate functions, search your codebase semantically, and refactor projects—with support for **multiple LLM providers** including local (Ollama) and cloud options with free tiers.
+A powerful CLI tool and **VSCode extension** that brings AI-powered code cognition to your development workflow. Review code, generate functions, search your codebase semantically, and refactor projects—with support for **multiple LLM providers** including local (Ollama) and cloud options with free tiers.
 
 ## ✨ Features
 
@@ -95,10 +97,67 @@ A powerful CLI tool that brings AI-powered code cognition to your terminal. Revi
 | 🔎 **Semantic Search** | Search your codebase using natural language queries |
 | 📝 **AI File Editing** | Edit files with natural language instructions |
 | 🔄 **Multi-File Refactor** | Refactor across multiple files at once |
-| ��️ **Symbol Renaming** | Rename functions, classes, variables across your project |
+| ✏️ **Symbol Renaming** | Rename functions, classes, variables across your project |
 | 💬 **Interactive Chat** | Chat with AI about your code |
 | 📊 **Codebase Indexing** | Create searchable semantic index with RAG |
 | 🌐 **Multi-Provider** | Support for 6 LLM providers (local & cloud) |
+| 🖥️ **VSCode Extension** | Full IDE integration with sidebar chat |
+
+## 🖥️ VSCode Extension
+
+<p align="center">
+  <img src="vscode-extension/images/icon.png" width="64" alt="Cognify VSCode">
+</p>
+
+The Cognify AI VSCode extension brings all the power of Cognify directly into your IDE with a beautiful sidebar chat interface.
+
+### Extension Features
+
+- ⚛️ **Sidebar Chat Panel** - Chat with AI directly in VSCode sidebar
+- 📎 **Add Context** - Include code from your editor in conversations
+- 🔍 **Code Review** - Review files with inline diagnostics
+- ✨ **Code Generation** - Generate code from descriptions
+- 💡 **Code Explanation** - Get explanations for selected code
+- ✏️ **AI Editing** - Edit code with natural language
+
+### Quick Actions
+- 📋 Review current file
+- 💡 Explain selected code
+- ✨ Suggest improvements
+- 🧪 Generate tests
+
+### Install VSCode Extension
+
+```bash
+# Build from source
+cd vscode-extension
+npm install
+npm run package
+
+# Install the extension
+code --install-extension cognify-ai-0.2.0.vsix
+```
+
+Or install via VSCode:
+1. Open VSCode
+2. Go to Extensions sidebar
+3. Click "..." → "Install from VSIX..."
+4. Select `cognify-ai-0.2.0.vsix`
+
+### Using the Extension
+
+1. Click the ⚛️ Cognify icon in the Activity Bar (left sidebar)
+2. Use quick actions or type your question
+3. Click "📎 Add Context" to include code from your editor
+4. Press Enter or click Send
+
+**Keyboard Shortcuts:**
+| Command | Mac | Windows/Linux |
+|---------|-----|---------------|
+| Review File | `Cmd+Shift+R` | `Ctrl+Shift+R` |
+| Generate Code | `Cmd+Shift+G` | `Ctrl+Shift+G` |
+| Explain Code | `Cmd+Shift+E` | `Ctrl+Shift+E` |
+| Open Chat | `Cmd+Shift+C` | `Ctrl+Shift+C` |
 
 ## 🤖 Supported Providers
 
@@ -153,87 +212,46 @@ pip install -e .
 
 ```bash
 # Check status and available providers
-ai-assist status
-ai-assist providers
-```
-
-## 🌐 Provider Management
-
-### List Available Providers
-```bash
-ai-assist providers
+cognify status
+cognify providers
 ```
-Shows all providers with their models, free tier status, and API key requirements.
-
-### Switch Providers
-```bash
-# Switch to Groq (fast cloud inference)
-ai-assist use-provider groq --test
 
-# Switch to Google with specific model
-ai-assist use-provider google --model gemini-1.5-pro --test
-
-# Use OpenRouter with free DeepSeek R1
-ai-assist use-provider openrouter --model deepseek/deepseek-r1:free --test
-```
-
-### Test Provider Connection
-```bash
-ai-assist test-provider
-ai-assist test-provider --provider groq --prompt "Hello world"
-```
-
-## 📖 Usage
+## 📖 CLI Usage
 
 ### Code Review
 ```bash
-ai-assist review path/to/file.py
-ai-assist review src/ --format json
+cognify review path/to/file.py
+cognify review src/ --format json
 ```
 
 ### Code Generation
 ```bash
-ai-assist generate "binary search function" --language python
-ai-assist generate "REST API client class" --mode class
-ai-assist generate "unit tests for calculator" --mode test
+cognify generate "binary search function" --language python
+cognify generate "REST API client class" --mode class
 ```
 
 ### Semantic Search
 ```bash
 # First, index your codebase
-ai-assist index .
+cognify index .
 
 # Then search
-ai-assist search "error handling"
-ai-assist search "database connection" -k 10
+cognify search "error handling"
 ```
 
 ### File Editing
 ```bash
-ai-assist edit config.py "add logging to all functions" --preview
-ai-assist edit utils.py "add type hints" --backup
-```
-
-### Multi-File Refactoring
-```bash
-ai-assist refactor "add docstrings to all functions" -p "src/**/*.py" --dry-run
-ai-assist refactor "convert print to logging" --pattern "**/*.py" --confirm
-```
-
-### Symbol Renaming
-```bash
-ai-assist rename old_function new_function --type function --dry-run
-ai-assist rename MyClass BetterClass --type class -p "src/**/*.py"
+cognify edit config.py "add logging to all functions" --preview
 ```
 
 ### Interactive Chat
 ```bash
-ai-assist chat
+cognify chat
 ```
 
 ### All Commands
 ```bash
-ai-assist --help
+cognify --help
 ```
 
 ## ⚙️ Configuration
@@ -242,32 +260,11 @@ Configuration is managed via `config.yaml`:
 
 ```yaml
 llm:
-  provider: "ollama"          # ollama, google, groq, cerebras, openrouter, openai
+  provider: "ollama"
   model: "deepseek-coder:6.7b"
-  base_url: "http://localhost:11434"  # For Ollama
+  base_url: "http://localhost:11434"
   temperature: 0.1
   max_tokens: 4096
-  timeout: 120
-
-review:
-  severity_levels: [critical, warning, suggestion]
-  categories: [bugs, security, performance, style]
-
-generation:
-  include_type_hints: true
-  include_docstrings: true
-
-retrieval:
-  embedding_model: "all-MiniLM-L6-v2"
-  chunk_size: 50
-
-editor:
-  create_backup: true
-  show_diff: true
-
-refactor:
-  max_files: 20
-  require_confirmation: true
 ```
 
 Or use environment variables:
@@ -281,100 +278,36 @@ export GROQ_API_KEY="your-key"
 
 ```
 cognify-ai/
-├── src/ai_code_assistant/
-│   ├── cli.py              # Command-line interface
-│   ├── config.py           # Configuration management
-│   ├── llm.py              # LLM integration
-│   ├── providers/          # Multi-provider support
-│   │   ├── base.py         # Provider base class
-│   │   ├── factory.py      # Provider factory
-│   │   ├── ollama.py       # Ollama (local)
-│   │   ├── google.py       # Google AI Studio
-│   │   ├── groq.py         # Groq
-│   │   ├── cerebras.py     # Cerebras
-│   │   ├── openrouter.py   # OpenRouter
-│   │   └── openai.py       # OpenAI
-│   ├── reviewer/           # Code review module
-│   ├── generator/          # Code generation module
-│   ├── retrieval/          # Semantic search & indexing (RAG)
-│   ├── editor/             # AI file editing
-│   ├── refactor/           # Multi-file refactoring
-│   ├── chat/               # Interactive chat
-│   └── utils/              # Utilities & formatters
-├── tests/                  # 144 unit tests
-├── docs/                   # Documentation
-├── config.yaml             # Configuration
-└── pyproject.toml          # Dependencies
+├── src/ai_code_assistant/    # Core Python package
+│   ├── cli.py                # Command-line interface
+│   ├── providers/            # LLM providers
+│   ├── reviewer/             # Code review
+│   ├── generator/            # Code generation
+│   ├── retrieval/            # Semantic search
+│   ├── editor/               # AI file editing
+│   └── chat/                 # Interactive chat
+├── vscode-extension/         # VSCode Extension
+│   ├── src/                  # Extension source
+│   ├── images/               # Icons
+│   └── package.json          # Extension manifest
+├── tests/                    # Unit tests
+├── docs/                     # Documentation
+└── config.yaml               # Configuration
 ```
 
 ## 🧪 Testing
 
 ```bash
-# Run all tests
 PYTHONPATH=src pytest tests/ -v
-
-# Run with coverage
-PYTHONPATH=src pytest tests/ --cov=ai_code_assistant
-```
-
-## 🛠️ Tech Stack
-
-| Component | Technology |
-|-----------|------------|
-| LLM Framework | LangChain |
-| Local LLM | Ollama |
-| Cloud LLMs | Google, Groq, OpenRouter, OpenAI |
-| Vector Database | ChromaDB |
-| Embeddings | Sentence Transformers |
-| CLI | Click + Rich |
-| Config | Pydantic |
-| Testing | Pytest |
-
-## 🐛 Troubleshooting
-
-**"Connection refused" error (Ollama)**
-```bash
-ollama serve  # Make sure Ollama is running
-```
-
-**API Key errors**
-```bash
-ai-assist providers  # Check which API keys are set
-export GROQ_API_KEY="your-key"  # Set the appropriate key
-```
-
-**Test provider connection**
-```bash
-ai-assist test-provider --provider groq
-```
-
-**Import errors**
-```bash
-pip install -e ".[dev]"
 ```
 
 ## 🤝 Contributing
 
 Contributions are welcome! Please feel free to submit a Pull Request.
 
-1. Fork the repository
-2. Create your feature branch (`git checkout -b feature/amazing-feature`)
-3. Commit your changes (`git commit -m 'Add amazing feature'`)
-4. Push to the branch (`git push origin feature/amazing-feature`)
-5. Open a Pull Request
-
 ## 📄 License
 
-This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
-
-## 🙏 Acknowledgments
-
-- [Ollama](https://ollama.ai) - Local LLM runtime
-- [LangChain](https://langchain.com) - LLM framework
-- [Google AI Studio](https://aistudio.google.com) - Gemini models
-- [Groq](https://groq.com) - Fast inference
-- [OpenRouter](https://openrouter.ai) - Multi-provider access
-- [ChromaDB](https://www.trychroma.com) - Vector database
+MIT License - see [LICENSE](LICENSE) for details.
 
 ---
 

{cognify_code-0.2.5.dist-info → cognify_code-0.2.6.dist-info}/RECORD

@@ -1,10 +1,10 @@
 ai_code_assistant/__init__.py,sha256=XnpG4h-2gW3cXseFvqQT_-XyOmVJtikVMrHUnmy8XKI,409
-ai_code_assistant/cli.py,sha256=Js8CyT_OjBymChhBeXKWrt2fK8KT0_2tfrTttgtsZSE,65497
+ai_code_assistant/cli.py,sha256=vrz1zt8JbNXflXVpcSr1ZF_SphBrAPTwDcn-qGR5vfU,67912
 ai_code_assistant/config.py,sha256=6sAufexwzfCu2JNWvt9KevS9k_gMcjj1TAnwuaO1ZFw,4727
 ai_code_assistant/llm.py,sha256=DfcWJf6zEAUsPSEZLdEmb9o6BQNf1Ja88nswjpy6cOw,4209
 ai_code_assistant/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 ai_code_assistant/agent/__init__.py,sha256=BcVe4Ebopv_J01ApnRl05oN5yOet5mEefBrQmdPsUj0,1284
-ai_code_assistant/agent/code_agent.py,sha256=-zO9cANnYIIxLa4h-vJaqbxakHkBoa7Tek7z2RdWJkk,39668
+ai_code_assistant/agent/code_agent.py,sha256=lykEvcbGl9V4RdKp2sCrP8-jlL-C7mIqvcAwpXHjcSY,28518
 ai_code_assistant/agent/code_generator.py,sha256=rAaziRU-mJ5NooERjR_Cd6_hwO0kuULw3Sp8Ca9kR48,13138
 ai_code_assistant/agent/code_reviewer.py,sha256=YiM7lRJhoN-vBnQb29jF-5nmE9ppL-OJffvx4ocTHEU,12066
 ai_code_assistant/agent/diff_engine.py,sha256=A5jszowc5VmWbdidpIW_QhswG_Hats3FYuemP8VoYv4,11018
@@ -27,7 +27,7 @@ ai_code_assistant/git/__init__.py,sha256=YgqmzneAnZyRrbazMqGoFSPIk5Yf5OTm2LXPbkQ
 ai_code_assistant/git/commit_generator.py,sha256=CzDH5ZPqEaXyPznBg8FgTz8wbV4adALUQD__kl8au6o,4135
 ai_code_assistant/git/manager.py,sha256=BYeYSz3yPpeJJESy2Zmu4MKEvJ5YAtw3HAmU6uba3nM,6815
 ai_code_assistant/providers/__init__.py,sha256=T8eLHOcjWvqNxLsD8uLmU2H1mJbGbZgUrUcrrVRcqPs,832
-ai_code_assistant/providers/base.py,sha256=6NJxzidnf5e_0hrbq5PxL4qsk9lGGU1Uzk4WDiWsZso,3969
+ai_code_assistant/providers/base.py,sha256=Ep0ZXT2u3_nGUKItAMjf4fxDh2kDRT5CXaZn9P1r6Ys,7279
 ai_code_assistant/providers/cerebras.py,sha256=PfjfFtkFOip5OquyOnxlSQowpy8uPWNRLA6y4m-iYio,3098
 ai_code_assistant/providers/factory.py,sha256=U2zH3HFDGhed2nPRpTyDqG4JcFNHvTvxw25NER2NEi0,4579
 ai_code_assistant/providers/google.py,sha256=nEHsAUiBhV9TjtJEwxkMWydtnWiouVtl_2MrcU8GQNI,3344
@@ -50,9 +50,9 @@ ai_code_assistant/reviewer/prompts.py,sha256=9RrHEBttS5ngxY2BNsUvqGC6-cTxco-kDPb
 ai_code_assistant/utils/__init__.py,sha256=3HO-1Bj4VvUtM7W1C3MKR4DzQ9Xc875QKSHHkHwuqVs,368
 ai_code_assistant/utils/file_handler.py,sha256=jPxvtI5dJxkpPjELgRJ11WXamtyKKmZANQ1fcfMVtiU,5239
 ai_code_assistant/utils/formatters.py,sha256=5El9ew9HS6JLBucBUxxcw4fO5nLpOucgNJrJj2NC3zw,8945
-cognify_code-0.2.5.dist-info/licenses/LICENSE,sha256=5yu_kWq2bK-XKhWo79Eykdg4Qf3O8V2Ys7cpOO7GyyE,1063
-cognify_code-0.2.5.dist-info/METADATA,sha256=nXVlyJL2PI0JSu0rLtttjNgPTrhL9ZAJ6EejVEttKEM,11862
-cognify_code-0.2.5.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
-cognify_code-0.2.5.dist-info/entry_points.txt,sha256=MrBnnWPHZVozqqKyTlnJO63YN2kE5yPWKlr2nnRFRks,94
-cognify_code-0.2.5.dist-info/top_level.txt,sha256=dD_r1x-oX0s1uspYY72kig4jfIsjh3oDKwOBCMYXqpo,18
-cognify_code-0.2.5.dist-info/RECORD,,
+cognify_code-0.2.6.dist-info/licenses/LICENSE,sha256=5yu_kWq2bK-XKhWo79Eykdg4Qf3O8V2Ys7cpOO7GyyE,1063
+cognify_code-0.2.6.dist-info/METADATA,sha256=JD70dP-OBqiN8_VPo0T1qp-Ifw1OsQLesAD4sruE7vk,10189
+cognify_code-0.2.6.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
+cognify_code-0.2.6.dist-info/entry_points.txt,sha256=MrBnnWPHZVozqqKyTlnJO63YN2kE5yPWKlr2nnRFRks,94
+cognify_code-0.2.6.dist-info/top_level.txt,sha256=dD_r1x-oX0s1uspYY72kig4jfIsjh3oDKwOBCMYXqpo,18
+cognify_code-0.2.6.dist-info/RECORD,,