vibecore 0.2.0a1__tar.gz → 0.3.0__tar.gz

{vibecore-0.2.0a1 → vibecore-0.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: vibecore
-Version: 0.2.0a1
+Version: 0.3.0
 Summary: Build your own AI-powered automation tools in the terminal with this extensible agent framework
 Project-URL: Homepage, https://github.com/serialx/vibecore
 Project-URL: Repository, https://github.com/serialx/vibecore
@@ -27,6 +27,9 @@ Classifier: Topic :: Terminals
 Classifier: Topic :: Text Processing :: Linguistic
 Classifier: Typing :: Typed
 Requires-Python: >=3.11
+Requires-Dist: ddgs>=9.5.4
+Requires-Dist: html2text>=2024.2.26
+Requires-Dist: httpx>=0.24.0
 Requires-Dist: litellm>=1.72.4
 Requires-Dist: openai-agents[litellm]>=0.2.2
 Requires-Dist: pydantic-settings>=2.10.1
@@ -65,6 +68,7 @@ Built on [Textual](https://textual.textualize.io/) and the [OpenAI Agents SDK](h
 
 ### Key Features
 
+- **Flow Mode (Experimental)** - Build structured agent-based applications with programmatic conversation control
 - **AI-Powered Chat Interface** - Interact with state-of-the-art language models through an intuitive terminal interface
 - **Rich Tool Integration** - Built-in tools for file operations, shell commands, Python execution, and task management
 - **MCP Support** - Connect to external tools and services via Model Context Protocol servers
@@ -79,6 +83,26 @@ Built on [Textual](https://textual.textualize.io/) and the [OpenAI Agents SDK](h
 ### Prerequisites
 
 - Python 3.11 or higher
+- (Optional) [uv](https://docs.astral.sh/uv/) for quick testing and better package management
+
+### Quick Test (No Installation)
+
+Try vibecore instantly without installing it:
+
+```bash
+# Install uv if you don't have it (optional)
+curl -LsSf https://astral.sh/uv/install.sh | sh
+
+# Configure your API key
+export ANTHROPIC_API_KEY="your-api-key-here"
+# or
+export OPENAI_API_KEY="your-api-key-here"
+
+# Run vibecore directly with uvx
+uvx vibecore
+```
+
+This will download and run vibecore in an isolated environment without affecting your system Python installation.
 
 ### Install from PyPI
 
@@ -136,6 +160,88 @@ Once vibecore is running, you can:
 - `/help` - Show help and keyboard shortcuts
 - `/clear` - Clear the current session and start a new one
 
+## Flow Mode (Experimental)
+
+Flow Mode is vibecore's **key differentiator** - it transforms the framework from a chat interface into a platform for building structured agent-based applications with programmatic conversation control.
+
+### What is Flow Mode?
+
+Flow Mode allows you to:
+- **Define custom conversation logic** that controls how agents process user input
+- **Build multi-step workflows** with defined sequences and decision points
+- **Orchestrate multiple agents** with handoffs and shared context
+- **Maintain conversation state** across interactions
+- **Create agent-based applications** rather than just chatbots
+
+### Example: Simple Flow
+
+```python
+import asyncio
+from agents import Agent, Runner
+from vibecore.flow import flow, UserInputFunc
+from vibecore.context import VibecoreContext
+
+# Define your agent with tools
+agent = Agent[VibecoreContext](
+    name="Assistant",
+    instructions="You are a helpful assistant",
+    tools=[...],  # Your tools here
+)
+
+# Define your conversation logic
+async def logic(app, ctx: VibecoreContext, user_input: UserInputFunc):
+    # Get user input programmatically
+    user_message = await user_input("What would you like to do?")
+    
+    # Process with agent
+    result = Runner.run_streamed(
+        agent,
+        input=user_message,
+        context=ctx,
+        session=app.session,
+    )
+    
+    # Handle the response
+    app.current_worker = app.handle_streamed_response(result)
+    await app.current_worker.wait()
+
+# Run the flow
+async def main():
+    await flow(agent, logic)
+
+if __name__ == "__main__":
+    asyncio.run(main())
+```
+
+### Example: Multi-Agent Customer Service
+
+Flow Mode shines when building complex multi-agent systems. See `examples/customer_service.py` for a complete implementation featuring:
+
+- **Triage Agent**: Routes requests to appropriate specialists
+- **FAQ Agent**: Handles frequently asked questions
+- **Booking Agent**: Manages seat reservations
+- **Agent Handoffs**: Seamless transitions between agents with context preservation
+- **Shared State**: Maintains customer information across the conversation
+
+### Key Components
+
+- **`flow()`**: Entry point that sets up the Vibecore app with your custom logic
+- **`logic()`**: Your async function that controls the conversation flow
+- **`UserInputFunc`**: Provides programmatic user input collection
+- **`VibecoreContext`**: Shared state across tools and agents
+- **Agent Handoffs**: Transfer control between specialized agents
+
+### Use Cases
+
+Flow Mode enables building:
+- **Customer service systems** with routing and escalation
+- **Guided workflows** for complex tasks
+- **Interactive tutorials** with step-by-step guidance
+- **Task automation** with human-in-the-loop controls
+- **Multi-stage data processing** pipelines
+
+The examples in the `examples/` directory are adapted from the official OpenAI Agents SDK with minimal modifications, demonstrating how easily you can build sophisticated agent applications with vibecore.
+
 ### Available Tools
 
 vibecore comes with powerful built-in tools:

{vibecore-0.2.0a1 → vibecore-0.3.0}/README.md

@@ -29,6 +29,7 @@ Built on [Textual](https://textual.textualize.io/) and the [OpenAI Agents SDK](h
 
 ### Key Features
 
+- **Flow Mode (Experimental)** - Build structured agent-based applications with programmatic conversation control
 - **AI-Powered Chat Interface** - Interact with state-of-the-art language models through an intuitive terminal interface
 - **Rich Tool Integration** - Built-in tools for file operations, shell commands, Python execution, and task management
 - **MCP Support** - Connect to external tools and services via Model Context Protocol servers
@@ -43,6 +44,26 @@ Built on [Textual](https://textual.textualize.io/) and the [OpenAI Agents SDK](h
 ### Prerequisites
 
 - Python 3.11 or higher
+- (Optional) [uv](https://docs.astral.sh/uv/) for quick testing and better package management
+
+### Quick Test (No Installation)
+
+Try vibecore instantly without installing it:
+
+```bash
+# Install uv if you don't have it (optional)
+curl -LsSf https://astral.sh/uv/install.sh | sh
+
+# Configure your API key
+export ANTHROPIC_API_KEY="your-api-key-here"
+# or
+export OPENAI_API_KEY="your-api-key-here"
+
+# Run vibecore directly with uvx
+uvx vibecore
+```
+
+This will download and run vibecore in an isolated environment without affecting your system Python installation.
 
 ### Install from PyPI
 
@@ -100,6 +121,88 @@ Once vibecore is running, you can:
 - `/help` - Show help and keyboard shortcuts
 - `/clear` - Clear the current session and start a new one
 
+## Flow Mode (Experimental)
+
+Flow Mode is vibecore's **key differentiator** - it transforms the framework from a chat interface into a platform for building structured agent-based applications with programmatic conversation control.
+
+### What is Flow Mode?
+
+Flow Mode allows you to:
+- **Define custom conversation logic** that controls how agents process user input
+- **Build multi-step workflows** with defined sequences and decision points
+- **Orchestrate multiple agents** with handoffs and shared context
+- **Maintain conversation state** across interactions
+- **Create agent-based applications** rather than just chatbots
+
+### Example: Simple Flow
+
+```python
+import asyncio
+from agents import Agent, Runner
+from vibecore.flow import flow, UserInputFunc
+from vibecore.context import VibecoreContext
+
+# Define your agent with tools
+agent = Agent[VibecoreContext](
+    name="Assistant",
+    instructions="You are a helpful assistant",
+    tools=[...],  # Your tools here
+)
+
+# Define your conversation logic
+async def logic(app, ctx: VibecoreContext, user_input: UserInputFunc):
+    # Get user input programmatically
+    user_message = await user_input("What would you like to do?")
+    
+    # Process with agent
+    result = Runner.run_streamed(
+        agent,
+        input=user_message,
+        context=ctx,
+        session=app.session,
+    )
+    
+    # Handle the response
+    app.current_worker = app.handle_streamed_response(result)
+    await app.current_worker.wait()
+
+# Run the flow
+async def main():
+    await flow(agent, logic)
+
+if __name__ == "__main__":
+    asyncio.run(main())
+```
+
+### Example: Multi-Agent Customer Service
+
+Flow Mode shines when building complex multi-agent systems. See `examples/customer_service.py` for a complete implementation featuring:
+
+- **Triage Agent**: Routes requests to appropriate specialists
+- **FAQ Agent**: Handles frequently asked questions
+- **Booking Agent**: Manages seat reservations
+- **Agent Handoffs**: Seamless transitions between agents with context preservation
+- **Shared State**: Maintains customer information across the conversation
+
+### Key Components
+
+- **`flow()`**: Entry point that sets up the Vibecore app with your custom logic
+- **`logic()`**: Your async function that controls the conversation flow
+- **`UserInputFunc`**: Provides programmatic user input collection
+- **`VibecoreContext`**: Shared state across tools and agents
+- **Agent Handoffs**: Transfer control between specialized agents
+
+### Use Cases
+
+Flow Mode enables building:
+- **Customer service systems** with routing and escalation
+- **Guided workflows** for complex tasks
+- **Interactive tutorials** with step-by-step guidance
+- **Task automation** with human-in-the-loop controls
+- **Multi-stage data processing** pipelines
+
+The examples in the `examples/` directory are adapted from the official OpenAI Agents SDK with minimal modifications, demonstrating how easily you can build sophisticated agent applications with vibecore.
+
 ### Available Tools
 
 vibecore comes with powerful built-in tools:

{vibecore-0.2.0a1 → vibecore-0.3.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "vibecore"
-version = "0.2.0a1"
+version = "0.3.0"
 description = "Build your own AI-powered automation tools in the terminal with this extensible agent framework"
 readme = "README.md"
 authors = [
@@ -44,6 +44,9 @@ classifiers = [
     "Typing :: Typed",
 ]
 dependencies = [
+    "ddgs>=9.5.4",
+    "html2text>=2024.2.26",
+    "httpx>=0.24.0",
     "litellm>=1.72.4",
     "openai-agents[litellm]>=0.2.2",
     "pydantic-settings>=2.10.1",
@@ -59,7 +62,7 @@ dependencies = [
 "Documentation" = "https://github.com/serialx/vibecore#readme"
 
 [project.scripts]
-vibecore = "vibecore.cli:main"
+vibecore = "vibecore.cli:cli_main"
 
 [[tool.uv.index]]
 name = "testpypi"

{vibecore-0.2.0a1 → vibecore-0.3.0}/src/vibecore/agents/default.py

@@ -1,8 +1,7 @@
 from typing import TYPE_CHECKING
 
-from agents import Agent, ModelSettings
+from agents import Agent
 from agents.extensions.handoff_prompt import prompt_with_handoff_instructions
-from openai.types import Reasoning
 
 from vibecore.context import VibecoreContext
 from vibecore.settings import settings
@@ -11,6 +10,8 @@ from vibecore.tools.python.tools import execute_python
 from vibecore.tools.shell.tools import bash, glob, grep, ls
 from vibecore.tools.task.tools import task
 from vibecore.tools.todo.tools import todo_read, todo_write
+from vibecore.tools.webfetch.tools import webfetch
+from vibecore.tools.websearch.tools import websearch
 
 from .prompts import COMMON_PROMPT
 
@@ -50,26 +51,20 @@ def create_default_agent(mcp_servers: list["MCPServer"] | None = None) -> Agent[
         grep,
         ls,
         task,
+        websearch,
+        webfetch,
     ]
     instructions = INSTRUCTIONS
 
     instructions = prompt_with_handoff_instructions(instructions)
 
-    # Configure reasoning based on settings
-    reasoning_config = Reasoning(summary="auto")
-    if settings.reasoning_effort is not None:
-        reasoning_config = Reasoning(effort=settings.reasoning_effort, summary="auto")
-
     return Agent[VibecoreContext](
         name="Vibecore Agent",
         handoff_description="A versatile general-purpose assistant",
         instructions=instructions,
         tools=tools,
         model=settings.model,
-        model_settings=ModelSettings(
-            include_usage=True,  # Ensure token usage is tracked in streaming mode
-            reasoning=reasoning_config,
-        ),
+        model_settings=settings.default_model_settings,
         handoffs=[],
         mcp_servers=mcp_servers or [],
     )

vibecore-0.2.0a1/src/vibecore/agents/task_agent.py → vibecore-0.3.0/src/vibecore/agents/task.py

@@ -1,8 +1,7 @@
 """Task-specific agent configuration for executing delegated tasks."""
 
-from agents import Agent, ModelSettings
+from agents import Agent
 from agents.extensions.handoff_prompt import prompt_with_handoff_instructions
-from openai.types import Reasoning
 
 from vibecore.context import VibecoreContext
 from vibecore.settings import settings
@@ -58,9 +57,6 @@ def create_task_agent(prompt: str) -> Agent[VibecoreContext]:
         instructions=instructions,
         tools=tools,
         model=settings.model,
-        model_settings=ModelSettings(
-            include_usage=True,  # Ensure token usage is tracked in streaming mode
-            reasoning=Reasoning(summary="auto"),
-        ),
+        model_settings=settings.default_model_settings,
         handoffs=[],
     )

vibecore-0.3.0/src/vibecore/auth/__init__.py

@@ -0,0 +1,15 @@
+"""Anthropic Pro/Max authentication module."""
+
+from vibecore.auth.config import ANTHROPIC_CONFIG
+from vibecore.auth.manager import AnthropicAuthManager
+from vibecore.auth.models import AnthropicAuth, ApiKeyCredentials, OAuthCredentials
+from vibecore.auth.storage import SecureAuthStorage
+
+__all__ = [
+    "ANTHROPIC_CONFIG",
+    "AnthropicAuth",
+    "AnthropicAuthManager",
+    "ApiKeyCredentials",
+    "OAuthCredentials",
+    "SecureAuthStorage",
+]

vibecore-0.3.0/src/vibecore/auth/config.py

@@ -0,0 +1,38 @@
+"""Configuration constants for Anthropic authentication."""
+
+from typing import Final
+
+
+class AnthropicConfig:
+    """Configuration for Anthropic OAuth and API."""
+
+    # OAuth Client Configuration
+    OAUTH_CLIENT_ID: Final[str] = "9d1c250a-e61b-44d9-88ed-5944d1962f5e"
+    OAUTH_SCOPES: Final[str] = "org:create_api_key user:profile user:inference"
+    OAUTH_REDIRECT_URI: Final[str] = "https://console.anthropic.com/oauth/code/callback"
+    OAUTH_RESPONSE_TYPE: Final[str] = "code"
+    OAUTH_CODE_CHALLENGE_METHOD: Final[str] = "S256"
+
+    # API Endpoints
+    CLAUDE_AI_AUTHORIZE: Final[str] = "https://claude.ai/oauth/authorize"
+    CONSOLE_AUTHORIZE: Final[str] = "https://console.anthropic.com/oauth/authorize"
+    TOKEN_EXCHANGE: Final[str] = "https://console.anthropic.com/v1/oauth/token"
+    API_BASE: Final[str] = "https://api.anthropic.com"
+    MESSAGES: Final[str] = "https://api.anthropic.com/v1/messages"
+
+    # Beta Headers (Critical for Claude Code spoofing)
+    BETA_OAUTH: Final[str] = "oauth-2025-04-20"
+    BETA_CLAUDE_CODE: Final[str] = "claude-code-20250219"  # CRITICAL: Identifies as Claude Code
+    BETA_INTERLEAVED_THINKING: Final[str] = "interleaved-thinking-2025-05-14"
+    BETA_FINE_GRAINED_STREAMING: Final[str] = "fine-grained-tool-streaming-2025-05-14"
+
+    # Token Management
+    TOKEN_REFRESH_BUFFER_SECONDS: Final[int] = 300  # Refresh 5 minutes before expiry
+    TOKEN_MAX_RETRY_ATTEMPTS: Final[int] = 3
+    TOKEN_RETRY_DELAY_MS: Final[int] = 1000
+
+    # Claude Code Identity
+    CLAUDE_CODE_IDENTITY: Final[str] = "You are Claude Code, Anthropic's official CLI for Claude."
+
+
+ANTHROPIC_CONFIG = AnthropicConfig()

vibecore-0.3.0/src/vibecore/auth/interceptor.py

@@ -0,0 +1,141 @@
+"""HTTP request interceptor for Claude Code spoofing."""
+
+from typing import Any
+
+import httpx
+from httpx import URL
+
+from vibecore.auth.config import ANTHROPIC_CONFIG
+from vibecore.auth.storage import SecureAuthStorage
+from vibecore.auth.token_manager import TokenRefreshManager
+
+
+class AnthropicRequestInterceptor:
+    """Intercepts and modifies Anthropic API requests."""
+
+    def __init__(self, storage: SecureAuthStorage):
+        """
+        Initialize request interceptor.
+
+        Args:
+            storage: Secure storage for credentials.
+        """
+        self.storage = storage
+        self.token_manager = TokenRefreshManager(storage)
+
+    async def intercept_request(self, url: str, headers: dict[str, str] | None = None, **kwargs: Any) -> dict[str, Any]:
+        """
+        Intercept and modify request for Anthropic API.
+
+        Args:
+            url: Request URL.
+            headers: Request headers.
+            **kwargs: Additional request parameters.
+
+        Returns:
+            Modified request parameters.
+        """
+        auth = await self.storage.load("anthropic")
+
+        if not auth:
+            raise ValueError("Not authenticated with Anthropic")
+
+        # Prepare headers
+        headers = {} if headers is None else headers.copy()
+
+        if auth.type == "oauth":  # OAuth auth
+            await self._configure_oauth_headers(headers)
+        else:  # API key auth
+            self._configure_api_key_headers(headers, auth.key)  # type: ignore
+
+        # Apply Claude Code spoofing headers
+        self._apply_claude_code_headers(headers)
+
+        # Return modified request parameters
+        return {**kwargs, "headers": headers}
+
+    async def _configure_oauth_headers(self, headers: dict[str, str]) -> None:
+        """Configure headers for OAuth authentication."""
+        # Get valid access token (handles refresh automatically)
+        access_token = await self.token_manager.get_valid_token()
+
+        # Remove any API key headers (OAuth takes precedence)
+        headers.pop("x-api-key", None)
+        headers.pop("X-Api-Key", None)
+        headers.pop("anthropic-api-key", None)
+
+        # Set OAuth bearer token
+        headers["Authorization"] = f"Bearer {access_token}"
+
+    def _configure_api_key_headers(self, headers: dict[str, str], api_key: str) -> None:
+        """Configure headers for API key authentication."""
+        # Remove OAuth headers if present
+        headers.pop("Authorization", None)
+
+        # Set API key
+        headers["x-api-key"] = api_key
+
+    def _apply_claude_code_headers(self, headers: dict[str, str]) -> None:
+        """Apply Claude Code spoofing headers."""
+        # Build beta features header
+        beta_features = [
+            ANTHROPIC_CONFIG.BETA_OAUTH,
+            ANTHROPIC_CONFIG.BETA_CLAUDE_CODE,  # Critical for spoofing
+            ANTHROPIC_CONFIG.BETA_INTERLEAVED_THINKING,
+            # ANTHROPIC_CONFIG.BETA_FINE_GRAINED_STREAMING,
+        ]
+
+        # Set the beta header (this is what makes Anthropic think we're Claude Code)
+        headers["anthropic-beta"] = ",".join(beta_features)
+
+        # Set additional headers that Claude Code uses
+        headers["anthropic-version"] = "2023-06-01"
+        headers.setdefault("accept", "application/json")
+
+        # Add Claude Code specific headers
+        headers["user-agent"] = "Claude-Code/1.0"
+        headers["x-client-id"] = ANTHROPIC_CONFIG.OAUTH_CLIENT_ID
+
+        # Ensure content-type is set for POST requests
+        headers.setdefault("content-type", "application/json")
+
+
+class GlobalAnthropicInterceptor:
+    """Global interceptor for automatic Anthropic request modification."""
+
+    def __init__(self, storage: SecureAuthStorage):
+        """
+        Initialize global interceptor.
+
+        Args:
+            storage: Secure storage for credentials.
+        """
+        self.interceptor = AnthropicRequestInterceptor(storage)
+        self.original_client_class = httpx.AsyncClient
+
+    def install(self) -> None:
+        """Install global request interception."""
+        interceptor = self.interceptor
+
+        class InterceptedAsyncClient(httpx.AsyncClient):
+            """Custom AsyncClient that intercepts Anthropic requests."""
+
+            async def request(self, method: str, url: URL | str, **kwargs: Any) -> httpx.Response:
+                """Override request method to intercept Anthropic API calls."""
+                # Convert URL to string if needed
+                url_str = str(url)
+
+                # Check if this is an Anthropic API request
+                if "anthropic.com" in url_str or "claude.ai" in url_str:
+                    # Intercept and modify request
+                    kwargs = await interceptor.intercept_request(url_str, **kwargs)
+
+                # Call original request method
+                return await super().request(method, url, **kwargs)
+
+        # Replace httpx.AsyncClient globally
+        httpx.AsyncClient = InterceptedAsyncClient  # type: ignore
+
+    def uninstall(self) -> None:
+        """Uninstall global request interception."""
+        httpx.AsyncClient = self.original_client_class  # type: ignore