hellochusquis 0.6.0__tar.gz

hellochusquis-0.6.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Arnau Moyano
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

hellochusquis-0.6.0/PKG-INFO

@@ -0,0 +1,151 @@
+Metadata-Version: 2.4
+Name: hellochusquis
+Version: 0.6.0
+Summary: A powerful terminal AI agent with multi-provider fallback, plugin system, and web interface
+Author: aminoy77
+License: MIT
+Project-URL: Homepage, https://github.com/aminoy77/HelloChusquis
+Keywords: ai,agent,cli,llm,automation
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: rich>=13.0.0
+Requires-Dist: pyyaml>=6.0
+Requires-Dist: httpx>=0.24.0
+Requires-Dist: fastapi>=0.100.0
+Requires-Dist: uvicorn>=0.23.0
+Dynamic: license-file
+
+# HelloChusquis
+
+A powerful terminal AI agent built in Python with web interface, multi-provider fallback, plugin system, persistent memory, and auto-learning.
+
+## Features
+
+- 🤖 **AI Agent** — chat in terminal or browser with full tool access
+- 🔄 **Multi-provider fallback** — automatically switches between providers when one fails
+- 🧩 **Plugin system** — install plugins with one command, build your own
+- 🧠 **Persistent memory** — remembers past conversations and learns from them
+- 📋 **Task planner** — detects complex tasks and executes them step by step
+- 🌐 **Web interface** — chat from your browser with sidebar showing status
+- 📄 **Real file creation** — generates real PDF and Word documents
+- 🔍 **Browser control** — search the web and extract content from pages
+
+## Supported Providers
+
+OpenRouter, Ollama Cloud, Anthropic Claude, OpenAI, Google Gemini, Groq, xAI (Grok), Perplexity, Qwen, MiniMax, Mistral, DeepSeek, Cohere, Together AI, Fireworks AI, Novita AI, and more.
+
+## Installation
+
+### Option 1 — Download ZIP
+
+1. Go to [github.com/aminoy77/HelloChusquis](https://github.com/aminoy77/HelloChusquis)
+2. Click **Code → Download ZIP**
+3. Unzip the file
+4. Open terminal in the unzipped folder
+5. Run:
+```bash
+pip install -e .
+```
+
+### Option 2 — Clone repo
+
+```bash
+git clone https://github.com/aminoy77/HelloChusquis.git
+cd HelloChusquis
+pip install -e .
+```
+
+## First Run
+
+```bash
+hellochusquis
+```
+
+First run will launch the setup wizard to configure your providers. It is recommended to add at least 2 providers for automatic fallback.
+
+Get free API keys at:
+- [openrouter.ai](https://openrouter.ai) — access 300+ models, many free
+- [console.groq.com](https://console.groq.com) — fastest free tier
+- [ollama.com](https://ollama.com) — cloud models
+
+## Usage
+
+### Terminal
+
+```bash
+hellochusquis          # Start chat in terminal
+hellochusquis web      # Start web interface (opens browser)
+```
+
+### Chat Commands
+
+| Command | Description |
+|---|---|
+| `/help` | Show available commands |
+| `/status` | Show provider status |
+| `/clear` | Clear conversation history |
+| `/plan <task>` | Force planning mode for a task |
+| `👍` or `+` | Positive feedback on last response |
+| `👎` or `-` | Negative feedback on last response |
+| `exit` | Exit and save memory |
+
+### Terminal Commands
+
+```bash
+hellochusquis install      # Install a plugin
+hellochusquis uninstall    # Remove a plugin
+hellochusquis plugins              # List installed plugins
+hellochusquis build                # Build a new plugin with AI
+hellochusquis learn                # Show learned patterns and rules
+hellochusquis web                  # Start web interface
+```
+
+## Plugins
+
+Install plugins from the public registry:
+
+```bash
+hellochusquis install weather
+hellochusquis install stocks
+hellochusquis install browser
+hellochusquis install pdf
+hellochusquis install docx
+hellochusquis install crypto
+hellochusquis install calculator
+hellochusquis install worldclock
+```
+
+Browse all available plugins at [github.com/aminoy77/Hellochusquis-plugins](https://github.com/aminoy77/HelloChusquis-plugins.git)
+
+### Build your own plugin
+
+```bash
+hellochusquis build
+```
+
+The AI will research the API, write the plugin code, test it, and ask if you want to submit it to the public registry.
+
+## Web Interface
+
+```bash
+hellochusquis web
+```
+
+Opens a web chat at `http://localhost:8000` with:
+- Chat interface with tool call visualization
+- Sidebar with provider status, installed plugins, memory stats, and learned patterns
+
+## How it works
+
+**Provider fallback** — if a provider hits rate limits or times out, HelloChusquis automatically switches to the next one. Providers that support tool calling are prioritized for tasks that need tools.
+
+**Task planner** — for complex tasks, HelloChusquis generates a step-by-step plan, shows it to you for confirmation, then executes each step automatically.
+
+**Memory** — at the end of each session, HelloChusquis summarizes the conversation and stores it. Next session it remembers what you discussed.
+
+**Auto-learning** — after each session, the agent analyzes what worked and what didn't, and improves its behavior over time.
+
+## License
+
+MIT

hellochusquis-0.6.0/README.md

@@ -0,0 +1,133 @@
+# HelloChusquis
+
+A powerful terminal AI agent built in Python with web interface, multi-provider fallback, plugin system, persistent memory, and auto-learning.
+
+## Features
+
+- 🤖 **AI Agent** — chat in terminal or browser with full tool access
+- 🔄 **Multi-provider fallback** — automatically switches between providers when one fails
+- 🧩 **Plugin system** — install plugins with one command, build your own
+- 🧠 **Persistent memory** — remembers past conversations and learns from them
+- 📋 **Task planner** — detects complex tasks and executes them step by step
+- 🌐 **Web interface** — chat from your browser with sidebar showing status
+- 📄 **Real file creation** — generates real PDF and Word documents
+- 🔍 **Browser control** — search the web and extract content from pages
+
+## Supported Providers
+
+OpenRouter, Ollama Cloud, Anthropic Claude, OpenAI, Google Gemini, Groq, xAI (Grok), Perplexity, Qwen, MiniMax, Mistral, DeepSeek, Cohere, Together AI, Fireworks AI, Novita AI, and more.
+
+## Installation
+
+### Option 1 — Download ZIP
+
+1. Go to [github.com/aminoy77/HelloChusquis](https://github.com/aminoy77/HelloChusquis)
+2. Click **Code → Download ZIP**
+3. Unzip the file
+4. Open terminal in the unzipped folder
+5. Run:
+```bash
+pip install -e .
+```
+
+### Option 2 — Clone repo
+
+```bash
+git clone https://github.com/aminoy77/HelloChusquis.git
+cd HelloChusquis
+pip install -e .
+```
+
+## First Run
+
+```bash
+hellochusquis
+```
+
+First run will launch the setup wizard to configure your providers. It is recommended to add at least 2 providers for automatic fallback.
+
+Get free API keys at:
+- [openrouter.ai](https://openrouter.ai) — access 300+ models, many free
+- [console.groq.com](https://console.groq.com) — fastest free tier
+- [ollama.com](https://ollama.com) — cloud models
+
+## Usage
+
+### Terminal
+
+```bash
+hellochusquis          # Start chat in terminal
+hellochusquis web      # Start web interface (opens browser)
+```
+
+### Chat Commands
+
+| Command | Description |
+|---|---|
+| `/help` | Show available commands |
+| `/status` | Show provider status |
+| `/clear` | Clear conversation history |
+| `/plan <task>` | Force planning mode for a task |
+| `👍` or `+` | Positive feedback on last response |
+| `👎` or `-` | Negative feedback on last response |
+| `exit` | Exit and save memory |
+
+### Terminal Commands
+
+```bash
+hellochusquis install      # Install a plugin
+hellochusquis uninstall    # Remove a plugin
+hellochusquis plugins              # List installed plugins
+hellochusquis build                # Build a new plugin with AI
+hellochusquis learn                # Show learned patterns and rules
+hellochusquis web                  # Start web interface
+```
+
+## Plugins
+
+Install plugins from the public registry:
+
+```bash
+hellochusquis install weather
+hellochusquis install stocks
+hellochusquis install browser
+hellochusquis install pdf
+hellochusquis install docx
+hellochusquis install crypto
+hellochusquis install calculator
+hellochusquis install worldclock
+```
+
+Browse all available plugins at [github.com/aminoy77/Hellochusquis-plugins](https://github.com/aminoy77/HelloChusquis-plugins.git)
+
+### Build your own plugin
+
+```bash
+hellochusquis build
+```
+
+The AI will research the API, write the plugin code, test it, and ask if you want to submit it to the public registry.
+
+## Web Interface
+
+```bash
+hellochusquis web
+```
+
+Opens a web chat at `http://localhost:8000` with:
+- Chat interface with tool call visualization
+- Sidebar with provider status, installed plugins, memory stats, and learned patterns
+
+## How it works
+
+**Provider fallback** — if a provider hits rate limits or times out, HelloChusquis automatically switches to the next one. Providers that support tool calling are prioritized for tasks that need tools.
+
+**Task planner** — for complex tasks, HelloChusquis generates a step-by-step plan, shows it to you for confirmation, then executes each step automatically.
+
+**Memory** — at the end of each session, HelloChusquis summarizes the conversation and stores it. Next session it remembers what you discussed.
+
+**Auto-learning** — after each session, the agent analyzes what worked and what didn't, and improves its behavior over time.
+
+## License
+
+MIT

hellochusquis-0.6.0/cli.py

@@ -0,0 +1,61 @@
+import sys
+import os
+sys.path.insert(0, os.path.dirname(__file__))
+
+
+def main():
+    args = sys.argv[1:]
+
+    if args and args[0] == "install" and len(args) > 1:
+        from core.plugins import install_plugin
+        install_plugin(args[1])
+        return
+
+    if args and args[0] == "uninstall" and len(args) > 1:
+        from core.plugins import uninstall_plugin
+        uninstall_plugin(args[1])
+        return
+
+    if args and args[0] == "plugins":
+        from core.plugins import list_plugins
+        list_plugins()
+        return
+
+    if args and args[0] == "build":
+        from core.setup import ensure_config
+        from core.provider import ProviderPool
+        ensure_config()
+        pool = ProviderPool()
+        from core.builder import build_plugin
+        build_plugin(pool)
+        return
+
+    if args and args[0] == "learn":
+        from core.setup import ensure_config
+        from core.learning import load_learnings
+        from rich.console import Console
+        import json
+        ensure_config()
+        console = Console()
+        learnings = load_learnings()
+        console.print("\n[bold]Current learnings:[/bold]")
+        console.print_json(json.dumps(learnings, indent=2, ensure_ascii=False))
+        return
+
+    if args and args[0] == "web":
+        import webbrowser
+        from rich.console import Console
+        console = Console()
+        console.print("[cyan]Starting HelloChusquis web interface...[/cyan]")
+        console.print("[dim]Open: http://localhost:8000[/dim]")
+        webbrowser.open("http://localhost:8000")
+        from web.server import start
+        start()
+        return
+
+    from main import main as run
+    run()
+
+
+if __name__ == "__main__":
+    main()

hellochusquis-0.6.0/core/agent.py

@@ -0,0 +1,189 @@
+import json
+import core.memory as memory
+import core.learning as learning
+from core.provider import ProviderPool
+from core.history import History
+from tools.shell import ShellTool
+from tools.files import FilesTool
+from tools.code import CodeTool
+from workspace.manager import WorkspaceManager
+from core.plugins import load_plugins
+
+
+TOOLS_SCHEMA = [
+    {
+        "type": "function",
+        "function": {
+            "name": "shell",
+            "description": "Execute a terminal command",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "command": {"type": "string", "description": "The command to run"}
+                },
+                "required": ["command"]
+            }
+        }
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "code",
+            "description": "Execute Python code",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "code": {"type": "string", "description": "Python code to execute"}
+                },
+                "required": ["code"]
+            }
+        }
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "files",
+            "description": "Read, write, delete, list files in the workspace",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "action": {
+                        "type": "string",
+                        "enum": ["read", "write", "delete", "list", "create_dir"]
+                    },
+                    "path": {"type": "string"},
+                    "content": {"type": "string", "description": "Required for write"}
+                },
+                "required": ["action", "path"]
+            }
+        }
+    }
+]
+
+
+class Agent:
+    def __init__(self, config: dict):
+        self.pool = ProviderPool()
+        self.history = History()
+        self.workspace = WorkspaceManager(config["settings"]["workspace_dirs"])
+        self.shell = ShellTool()
+        self.files = FilesTool(config["settings"]["workspace_dirs"])
+        self.code = CodeTool()
+        self.system_prompt = config["agent"]["system_prompt"]
+        self.workspace_dirs = config["settings"]["workspace_dirs"]
+
+        # Cargar memoria de sesiones anteriores
+        summary = memory.load_summary()
+        if summary:
+            self.system_prompt += f"\n\nWhat you remember from past sessions:\n{summary}"
+
+        # Cargar learnings
+        learnings = learning.load_learnings()
+        learning_prompt = learning.build_learning_prompt(learnings)
+        if learning_prompt:
+            self.system_prompt += f"\n\n{learning_prompt}"
+
+        # Cargar plugins
+        self.plugins = load_plugins()
+        for plugin in self.plugins:
+            TOOLS_SCHEMA.append(plugin["schema"])
+
+        # Informar al agente qué plugins tiene disponibles
+        if self.plugins:
+            plugin_names = ", ".join(p["name"] for p in self.plugins)
+            self.system_prompt += f"\n\nInstalled plugins available as tools: {plugin_names}. Use them directly without asking the user to install anything."
+
+    def _dispatch_tool(self, name: str, args: dict):
+        if name == "shell":
+            return self.shell.run(**args)
+        elif name == "code":
+            return self.code.run(**args)
+        elif name == "files":
+            path = args.get("path", "")
+            if not self.workspace.is_allowed(path):
+                granted = self.workspace.request_access(path)
+                if not granted:
+                    from tools.base import ToolResult
+                    return ToolResult(success=False, output="", error="Access denied by user")
+                self.files.allow_dir(path)
+            return self.files.run(**args)
+
+        for plugin in self.plugins:
+            if plugin["name"] == name:
+                try:
+                    result_text = plugin["run"](**args)
+                    from tools.base import ToolResult
+                    return ToolResult(success=True, output=str(result_text))
+                except Exception as e:
+                    from tools.base import ToolResult
+                    return ToolResult(success=False, output="", error=str(e))
+
+        from tools.base import ToolResult
+        return ToolResult(success=False, output="", error=f"Unknown tool: {name}")
+
+    def run(self, user_input: str) -> str:
+        self.history.add("user", user_input)
+
+        messages = [
+            {"role": "system", "content": (
+                self.system_prompt +
+                f"\n\nWorkspace directories: {', '.join(self.workspace_dirs)}. "
+                "Always use absolute paths when calling file tools. "
+                "Only use tools when strictly necessary. Never use shell or code tools just to print or display text."
+            )},
+            *self.history.get()
+        ]
+
+        while True:
+            response = self.pool.chat_with_retry(messages, tools=TOOLS_SCHEMA)
+            message = response["choices"][0]["message"]
+
+            if not message.get("tool_calls"):
+                content = message.get("content", "")
+                self.history.add("assistant", content)
+                return content
+
+            messages.append(message)
+
+            for tc in message["tool_calls"]:
+                tool_name = tc["function"]["name"]
+                tool_args = json.loads(tc["function"]["arguments"])
+
+                from ui.terminal import print_tool_call, print_tool_result
+                print_tool_call(tool_name, tool_args)
+                result = self._dispatch_tool(tool_name, tool_args)
+                print_tool_result(result.success, result.output)
+
+                messages.append({
+                    "role": "tool",
+                    "tool_call_id": tc["id"],
+                    "content": result.output if result.success else f"ERROR: {result.error}"
+                })
+
+    def summarize_and_save(self, retention_days: int = 30):
+        messages = self.history.get()
+        if not messages:
+            return
+
+        memory.save_session(messages)
+
+        summary_prompt = (
+            "Summarize the key facts, tasks completed, and important context "
+            "from this conversation in 3-5 bullet points. Be very concise. "
+            "This will be injected into future sessions as memory.\n\n"
+            + "\n".join(f"{m['role']}: {m['content'][:200]}" for m in messages)
+        )
+
+        try:
+            response = self.pool.chat_with_retry([
+                {"role": "user", "content": summary_prompt}
+            ])
+            summary = response["choices"][0]["message"]["content"]
+            memory.save_summary(summary)
+        except Exception:
+            pass
+
+        # Analizar y aprender de la sesión
+        learning.analyze_and_learn(messages, self.pool)
+
+        memory.cleanup_old_sessions(retention_days)