kader 2.7.2__tar.gz → 2.7.4__tar.gz

{kader-2.7.2 → kader-2.7.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kader
-Version: 2.7.2
+Version: 2.7.4
 Summary: kader coding agent
 License-File: LICENSE
 Requires-Python: >=3.11
@@ -33,6 +33,7 @@ Kader is an intelligent coding agent designed to assist with software developmen
 
 - 🤖 **AI-powered Code Assistance** - Support for multiple LLM providers:
   - **Ollama**: Local LLM execution for privacy and speed.
+  - **Ollama Cloud**: Cloud-based models via [ollama.com](https://ollama.com).
   - **Google Gemini**: Cloud-based powerful models via the Google GenAI SDK.
   - **Anthropic**: High-quality Claude models via the Anthropic SDK.
 - 🖥️ **Interactive CLI** - Modern terminal interface built with Rich & prompt_toolkit:
@@ -125,7 +126,7 @@ When the kader module is imported for the first time, it automatically creates a
 ### Environment Variables
 
 The application automatically loads environment variables from `~/.kader/.env`:
-- `OLLAMA_API_KEY`: API key for Ollama service (if applicable).
+- `OLLAMA_API_KEY`: API key for Ollama Cloud (for cloud models at ollama.com). Get your key from https://ollama.com/settings
 - `GOOGLE_API_KEY`: API key for Google Gemini (required for Google Provider).
 - `ANTHROPIC_API_KEY`: API key for Anthropic Claude (required for Anthropic Provider).
 - Additional variables can be added to the `.env` file and will be automatically loaded.
@@ -143,7 +144,7 @@ Kader stores data in `~/.kader/`:
 | Command | Description |
 |---------|-------------|
 | `/help` | Show command reference |
-| `/models` | Show available models (Ollama, Google & Anthropic) |
+| `/models` | Show available models (Ollama local & cloud, Google & Anthropic) |
 | `/clear` | Clear conversation |
 | `/save` | Save current session |
 | `/load <id>` | Load a saved session |
@@ -211,6 +212,7 @@ Kader provides a robust agent architecture:
 
 Kader supports multiple backends:
 - **OllamaProvider**: Connects to locally running Ollama instances.
+- **OllamaProvider (Cloud)**: Connects to cloud models at ollama.com (requires OLLAMA_API_KEY).
 - **GoogleProvider**: High-performance access to Gemini models.
 - **AnthropicProvider**: Full support for Claude models.
 

{kader-2.7.2 → kader-2.7.4}/README.md

@@ -6,6 +6,7 @@ Kader is an intelligent coding agent designed to assist with software developmen
 
 - 🤖 **AI-powered Code Assistance** - Support for multiple LLM providers:
   - **Ollama**: Local LLM execution for privacy and speed.
+  - **Ollama Cloud**: Cloud-based models via [ollama.com](https://ollama.com).
   - **Google Gemini**: Cloud-based powerful models via the Google GenAI SDK.
   - **Anthropic**: High-quality Claude models via the Anthropic SDK.
 - 🖥️ **Interactive CLI** - Modern terminal interface built with Rich & prompt_toolkit:
@@ -98,7 +99,7 @@ When the kader module is imported for the first time, it automatically creates a
 ### Environment Variables
 
 The application automatically loads environment variables from `~/.kader/.env`:
-- `OLLAMA_API_KEY`: API key for Ollama service (if applicable).
+- `OLLAMA_API_KEY`: API key for Ollama Cloud (for cloud models at ollama.com). Get your key from https://ollama.com/settings
 - `GOOGLE_API_KEY`: API key for Google Gemini (required for Google Provider).
 - `ANTHROPIC_API_KEY`: API key for Anthropic Claude (required for Anthropic Provider).
 - Additional variables can be added to the `.env` file and will be automatically loaded.
@@ -116,7 +117,7 @@ Kader stores data in `~/.kader/`:
 | Command | Description |
 |---------|-------------|
 | `/help` | Show command reference |
-| `/models` | Show available models (Ollama, Google & Anthropic) |
+| `/models` | Show available models (Ollama local & cloud, Google & Anthropic) |
 | `/clear` | Clear conversation |
 | `/save` | Save current session |
 | `/load <id>` | Load a saved session |
@@ -184,6 +185,7 @@ Kader provides a robust agent architecture:
 
 Kader supports multiple backends:
 - **OllamaProvider**: Connects to locally running Ollama instances.
+- **OllamaProvider (Cloud)**: Connects to cloud models at ollama.com (requires OLLAMA_API_KEY).
 - **GoogleProvider**: High-performance access to Gemini models.
 - **AnthropicProvider**: Full support for Claude models.
 

{kader-2.7.2 → kader-2.7.4}/cli/README.md

@@ -16,7 +16,7 @@ A modern terminal-based AI coding assistant built with [Rich](https://github.com
 
 ## Prerequisites
 
-- [Ollama](https://ollama.ai/) running locally (for local models)
+- [Ollama](https://ollama.ai/) running locally (for local models) or [Ollama Cloud](https://ollama.com) (for cloud models)
 - Python 3.11 or higher
 - [uv](https://docs.astral.sh/uv/) package manager (recommended) or [pip](https://pypi.org/project/pip/)
 - API keys for cloud providers (optional, based on model selection)
@@ -207,7 +207,8 @@ DEFAULT_MODEL = "minimax-m2.5:cloud"  # Default model
 
 | Provider | Format | Example |
 |----------|--------|---------|
-| Ollama | `ollama:model` | `ollama:llama3` |
+| Ollama (local) | `ollama:model` | `ollama:llama3` |
+| Ollama (cloud) | `ollama:model:cloud` | `ollama:minimax-m2.5:cloud` |
 | Google Gemini | `google:model` | `google:gemini-2.5-flash` |
 | Mistral | `mistral:model` | `mistral:small-3.1` |
 | Anthropic | `anthropic:model` | `anthropic:claude-3.5-sonnet` |
@@ -223,10 +224,16 @@ DEFAULT_MODEL = "minimax-m2.5:cloud"  # Default model
 Set API keys for cloud providers:
 
 ```bash
+# Ollama Cloud (get from https://ollama.com/settings)
+export OLLAMA_API_KEY="your-ollama-api-key"
+
+# Google Gemini
 export GOOGLE_API_KEY="your-google-api-key"
+
+# Other providers...
 export ANTHROPIC_API_KEY="your-anthropic-api-key"
 export MISTRAL_API_KEY="your-mistral-api-key"
-export OPENAI_API_KEY="your-openai-api-key"
+export OPENAI_API_KEY="your-openapi-key"
 export MOONSHOT_API_KEY="your-kimi-api-key"
 export ZAI_API_KEY="your-glm-api-key"
 export OPENROUTER_API_KEY="your-openrouter-api-key"

{kader-2.7.2 → kader-2.7.4}/cli/app.py

@@ -612,139 +612,28 @@ class KaderApp:
         self.console.print(f"\n  [kader.orange]\\[>] Executing:[/kader.orange] `{cmd}`")
 
         try:
-            output = await self._run_terminal_command_pty(cmd)
+            output = await self._run_terminal_command_direct(cmd)
 
             if not output:
-                self.console.print(
-                    "  [dim]Command executed successfully with no output.[/dim]"
+                output = "Command executed successfully with no output."
+
+            self.console.print()
+            self.console.print(
+                Panel(
+                    output,
+                    title="[kader.orange]Terminal Output[/kader.orange]",
+                    border_style="dark_orange",
+                    padding=(0, 1),
                 )
+            )
 
         except Exception as e:
             self.console.print(
                 rf"  [kader.red]\[-][/kader.red] Error executing command: {e}"
             )
 
-    async def _run_terminal_command_pty(self, command: str) -> str:
-        """Run a terminal command using PTY for interactive support."""
-        import platform
-
-        system = platform.system().lower()
-
-        if system == "windows":
-            return await self._run_terminal_command_winpty(command)
-        else:
-            return await self._run_terminal_command_unix_pty(command)
-
-    async def _run_terminal_command_unix_pty(self, command: str) -> str:
-        """Run a terminal command using PTY on Unix."""
-        import os
-        import pty
-
-        master_fd, slave_fd = pty.openpty()
-
-        env = os.environ.copy()
-        env["TERM"] = "xterm-256color"
-
-        shell = "/bin/bash"
-
-        try:
-            process = await asyncio.create_subprocess_exec(
-                shell,
-                "-c",
-                command,
-                stdin=slave_fd,
-                stdout=asyncio.subprocess.PIPE,
-                stderr=asyncio.subprocess.STDOUT,
-                env=env,
-            )
-        finally:
-            os.close(slave_fd)
-            os.close(master_fd)
-
-        output_parts = []
-        stdout = process.stdout
-        assert stdout is not None
-
-        while True:
-            line = await asyncio.wait_for(stdout.readline(), timeout=0.1)
-            if line:
-                text = line.decode("utf-8", errors="replace")
-                output_parts.append(text)
-                self.console.print(text, end="")
-            elif process.returncode is not None:
-                break
-
-        await process.wait()
-        return "".join(output_parts).strip()
-
-    async def _run_terminal_command_winpty(self, command: str) -> str:
-        """Run a terminal command using pywinpty on Windows."""
-
-        try:
-            import pywinpty
-        except ImportError:
-            return await self._run_terminal_command_direct(command)
-
-        command_lower = command.lower().strip()
-        is_powershell = command_lower.startswith("pwsh") or command_lower.startswith(
-            "powershell"
-        )
-
-        if is_powershell:
-            shell = "powershell.exe"
-            shell_args = ["-Command", command]
-        else:
-            shell = "cmd.exe"
-            shell_args = ["/c", command]
-
-        try:
-            pty_obj = pywinpty.PTY(width=80, height=24, visible=False)
-        except Exception:
-            return await self._run_terminal_command_direct(command)
-
-        try:
-            process = pty_obj.spawn(shell, shell_args)
-        except Exception:
-            return await self._run_terminal_command_direct(command)
-
-        output_parts = []
-        max_attempts = 100
-        attempts = 0
-
-        while attempts < max_attempts:
-            try:
-                data = process.read(blocking=False)
-                if data:
-                    output_parts.append(data)
-                    self.console.print(data, end="")
-                    attempts = 0
-            except Exception:
-                pass
-
-            if not process.isalive():
-                await asyncio.sleep(0.1)
-                try:
-                    data = process.read(blocking=False)
-                    if data:
-                        output_parts.append(data)
-                        self.console.print(data, end="")
-                except Exception:
-                    pass
-                if not process.isalive():
-                    break
-
-            attempts += 1
-            await asyncio.sleep(0.05)
-
-        result = "".join(output_parts).strip()
-
-        if not result:
-            return await self._run_terminal_command_direct(command)
-
-        return result
-
     async def _run_terminal_command_direct(self, command: str) -> str:
-        """Run a terminal command directly without PTY (fallback)."""
+        """Run a terminal command directly without PTY."""
 
         try:
             process = await asyncio.create_subprocess_shell(

{kader-2.7.2 → kader-2.7.4}/cli/llm_factory.py

@@ -16,6 +16,7 @@ from kader.providers import (
     OpenAIProviderConfig,
 )
 from kader.providers.base import BaseLLMProvider, ModelConfig
+from kader.providers.ollama import DEFAULT_CLOUD_HOST, DEFAULT_LOCAL_HOST
 
 
 class LLMProviderFactory:
@@ -168,6 +169,23 @@ class LLMProviderFactory:
                 default_config=config,
             )
 
+        # Handle Ollama provider (needs api_key for cloud models)
+        if provider_name == "ollama":
+            ollama_api_key = os.environ.get("OLLAMA_API_KEY")
+
+            # Strip :cloud suffix if present (we add it for display purposes)
+            actual_model_name = model_name.replace(":cloud", "")
+
+            # Use cloud host if API key is available
+            host = DEFAULT_CLOUD_HOST if ollama_api_key else None
+
+            return OllamaProvider(
+                model=actual_model_name,
+                host=host,
+                api_key=ollama_api_key,
+                default_config=config,
+            )
+
         return provider_class(model=model_name, default_config=config)
 
     @classmethod
@@ -226,13 +244,29 @@ class LLMProviderFactory:
         """
         models: dict[str, list[str]] = {}
 
-        # Get Ollama models
+        # Get Ollama models (local first)
         try:
-            ollama_models = OllamaProvider.get_supported_models()
+            ollama_models = OllamaProvider.get_supported_models(host=DEFAULT_LOCAL_HOST)
             models["ollama"] = [f"ollama:{m}" for m in ollama_models]
         except Exception:
             models["ollama"] = []
 
+        # Also try Ollama Cloud if API key is available
+        ollama_api_key = os.environ.get("OLLAMA_API_KEY")
+        if ollama_api_key:
+            try:
+                cloud_models = OllamaProvider.get_supported_models(
+                    host=DEFAULT_CLOUD_HOST, api_key=ollama_api_key
+                )
+                existing = set(models.get("ollama", []))
+                for m in cloud_models:
+                    # Add :cloud suffix to distinguish from local models
+                    prefixed = f"ollama:{m}:cloud"
+                    if prefixed.replace(":cloud", "") not in existing:
+                        models["ollama"].append(prefixed)
+            except Exception:
+                pass  # Cloud failed, but local models still available
+
         # Get Google models
         try:
             google_models = GoogleProvider.get_supported_models()
@@ -339,6 +373,23 @@ class LLMProviderFactory:
         # Try to get models to verify provider is working
         try:
             provider_class = cls.PROVIDERS[provider_name]
+
+            # Special handling for Ollama: check both local and cloud
+            if provider_name == "ollama":
+                local_models = provider_class.get_supported_models(
+                    host=DEFAULT_LOCAL_HOST
+                )
+                if local_models:
+                    return True
+                # Also try cloud if API key is available
+                ollama_api_key = os.environ.get("OLLAMA_API_KEY")
+                if ollama_api_key:
+                    cloud_models = provider_class.get_supported_models(
+                        host=DEFAULT_CLOUD_HOST, api_key=ollama_api_key
+                    )
+                    return len(cloud_models) > 0
+                return False
+
             models = provider_class.get_supported_models()
             return len(models) > 0
         except Exception:

{kader-2.7.2 → kader-2.7.4}/docs/cli/index.md

@@ -158,7 +158,8 @@ The model selection interface allows you to:
 
 | Provider | Format | Example |
 |----------|--------|---------|
-| Ollama | `ollama:model` | `ollama:llama3` |
+| Ollama (local) | `ollama:model` | `ollama:llama3` |
+| Ollama (cloud) | `ollama:model:cloud` | `ollama:minimax-m2.5:cloud` |
 | Google Gemini | `google:model` | `google:gemini-2.5-flash` |
 | Mistral | `mistral:model` | `mistral:small-3.1` |
 | Anthropic | `anthropic:model` | `anthropic:claude-3.5-sonnet` |
@@ -172,7 +173,13 @@ The model selection interface allows you to:
 ### Setting API Keys
 
 ```bash
+# Ollama Cloud (get from https://ollama.com/settings)
+export OLLAMA_API_KEY="your-ollama-api-key"
+
+# Google Gemini
 export GOOGLE_API_KEY="your-google-api-key"
+
+# Other providers...
 export ANTHROPIC_API_KEY="your-anthropic-api-key"
 export MISTRAL_API_KEY="your-mistral-api-key"
 export OPENAI_API_KEY="your-openai-api-key"

{kader-2.7.2 → kader-2.7.4}/docs/configuration.md

@@ -7,6 +7,7 @@ Kader can be configured through environment variables, YAML files, and the `.kad
 | Variable | Description | Required |
 |----------|-------------|----------|
 | `KADER_DIR` | Kader config directory (default: `~/.kader`) | No |
+| `OLLAMA_API_KEY` | Ollama Cloud API key (get from https://ollama.com/settings) | For Ollama Cloud |
 | `GEMINI_API_KEY` | Google Gemini API key | For Google Provider |
 | `MISTRAL_API_KEY` | Mistral API key | For Mistral Provider |
 | `ANTHROPIC_API_KEY` | Anthropic API key | For Anthropic Provider |
@@ -57,7 +58,7 @@ agent = BaseAgent.from_yaml("agent.yaml")
 
 ## Provider Configuration Format
 
-### Ollama
+### Ollama (Local)
 
 ```yaml
 provider:
@@ -67,6 +68,16 @@ provider:
   timeout: 120
 ```
 
+### Ollama (Cloud)
+
+```yaml
+provider:
+  provider: ollama
+  model: minimax-m2.5
+  base_url: "https://ollama.com"
+  # api_key: "your-ollama-api-key"  # Or set OLLAMA_API_KEY env var
+```
+
 ### Google Gemini
 
 ```yaml

{kader-2.7.2 → kader-2.7.4}/docs/guide.md

@@ -54,6 +54,9 @@ ollama pull qwen2.5
 Create a `.env` file in `~/.kader/.env`:
 
 ```bash
+# Ollama Cloud (get from https://ollama.com/settings)
+OLLAMA_API_KEY='your-api-key'
+
 # Google Gemini
 GEMINI_API_KEY='your-api-key'
 

{kader-2.7.2 → kader-2.7.4}/kader/README.md

@@ -41,6 +41,9 @@ pip install -e .
 For cloud providers, create a `.env` file in `~/.kader/.env`:
 
 ```bash
+# Ollama Cloud (get from https://ollama.com/settings)
+OLLAMA_API_KEY='your-api-key'
+
 # Google Gemini
 GEMINI_API_KEY='your-api-key'
 
@@ -66,6 +69,14 @@ ollama pull llama3.2
 ollama pull qwen2.5
 ```
 
+For cloud models via Ollama Cloud, get an API key from https://ollama.com/settings and set it in your environment:
+
+```bash
+export OLLAMA_API_KEY="your-api-key"
+```
+
+Then use cloud models with the `OllamaProvider` by specifying the cloud host:
+
 ## Core Concepts
 
 ### 1. Providers
@@ -302,10 +313,22 @@ For local LLM inference. Best for privacy, speed, and offline capability.
 ```python
 from kader.providers import OllamaProvider, Message
 
+# Local model (default - connects to localhost:11434)
 provider = OllamaProvider(
     model="llama3.2",          # Model name
-    base_url="http://localhost:11434",  # Custom endpoint
-    timeout=120,               # Request timeout
+)
+
+# Cloud model (requires OLLAMA_API_KEY environment variable)
+provider = OllamaProvider(
+    model="minimax-m2.5",      # Cloud model name
+    host="https://ollama.com",  # Cloud endpoint
+    api_key="your-api-key",     # Or set OLLAMA_API_KEY env var
+)
+
+# Custom local endpoint
+provider = OllamaProvider(
+    model="llama3.2",
+    host="http://localhost:11434",
 )
 
 # Basic invocation
@@ -907,6 +930,7 @@ python ollama_example.py
 | Variable | Description |
 |----------|-------------|
 | `KADER_DIR` | Kader config directory (default: `~/.kader`) |
+| `OLLAMA_API_KEY` | Ollama Cloud API key (for cloud models at ollama.com) |
 | `GEMINI_API_KEY` | Google Gemini API key |
 | `MISTRAL_API_KEY` | Mistral API key |
 | `OPENAI_API_KEY` | OpenAI API key |

{kader-2.7.2 → kader-2.7.4}/kader/providers/ollama.py

@@ -1,9 +1,11 @@
 """
 Ollama LLM Provider implementation.
 
-Provides synchronous and asynchronous access to Ollama models.
+Provides synchronous and asynchronous access to Ollama models,
+including both local and cloud (ollama.com) models.
 """
 
+import os
 from typing import AsyncIterator, Iterator
 
 from ollama import AsyncClient, Client
@@ -20,6 +22,9 @@ from .base import (
     Usage,
 )
 
+DEFAULT_LOCAL_HOST = "http://localhost:11434"
+DEFAULT_CLOUD_HOST = "https://ollama.com"
+
 
 class OllamaProvider(BaseLLMProvider):
     """
@@ -38,6 +43,7 @@ class OllamaProvider(BaseLLMProvider):
         self,
         model: str,
         host: str | None = None,
+        api_key: str | None = None,
         default_config: ModelConfig | None = None,
     ) -> None:
         """
@@ -46,12 +52,23 @@ class OllamaProvider(BaseLLMProvider):
         Args:
             model: The Ollama model identifier (e.g., "llama3.2", "gpt-oss:120b-cloud")
             host: Optional Ollama server host (default: http://localhost:11434)
+            api_key: Optional API key for Ollama Cloud (ollama.com)
             default_config: Default configuration for all requests
         """
         super().__init__(model=model, default_config=default_config)
         self._host = host
-        self._client = Client(host=host) if host else Client()
-        self._async_client = AsyncClient(host=host) if host else AsyncClient()
+        self._api_key = api_key or os.environ.get("OLLAMA_API_KEY")
+
+        headers = None
+        if self._api_key:
+            headers = {"Authorization": f"Bearer {self._api_key}"}
+
+        if host:
+            self._client = Client(host=host, headers=headers)
+            self._async_client = AsyncClient(host=host, headers=headers)
+        else:
+            self._client = Client()
+            self._async_client = AsyncClient()
 
     def _convert_messages(self, messages: list[Message]) -> list[dict]:
         """Convert Message objects to Ollama format."""
@@ -416,18 +433,28 @@ class OllamaProvider(BaseLLMProvider):
             return None
 
     @classmethod
-    def get_supported_models(cls, host: str | None = None) -> list[str]:
+    def get_supported_models(
+        cls, host: str | None = None, api_key: str | None = None
+    ) -> list[str]:
         """
         Get list of models available on the Ollama server.
 
         Args:
-            host: Optional Ollama server host
+            host: Optional Ollama server host (e.g., "http://localhost:11434" or "https://ollama.com")
+            api_key: Optional API key for Ollama Cloud
 
         Returns:
             List of available model names
         """
+        api_key = api_key or os.environ.get("OLLAMA_API_KEY")
+        headers = None
+        if api_key:
+            headers = {"Authorization": f"Bearer {api_key}"}
+
         try:
-            client = Client(host=host) if host else Client()
+            client = (
+                Client(host=host, headers=headers) if host else Client(headers=headers)
+            )
             response = client.list()
             models = [model.model for model in response.models]
             models_config = {}
@@ -444,4 +471,4 @@ class OllamaProvider(BaseLLMProvider):
 
     def list_models(self) -> list[str]:
         """List all available models on the Ollama server."""
-        return self.get_supported_models(self._host)
+        return self.get_supported_models(self._host, self._api_key)

{kader-2.7.2 → kader-2.7.4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "kader"
-version = "2.7.2"
+version = "2.7.4"
 description = "kader coding agent"
 readme = "README.md"
 requires-python = ">=3.11"

{kader-2.7.2 → kader-2.7.4}/uv.lock

@@ -629,7 +629,7 @@ wheels = [
 
 [[package]]
 name = "kader"
-version = "2.7.2"
+version = "2.7.4"
 source = { editable = "." }
 dependencies = [
     { name = "aiofiles" },