aru-code 0.1.0__tar.gz → 0.3.0__tar.gz

{aru_code-0.1.0 → aru_code-0.3.0}/LICENSE

@@ -1,21 +1,21 @@
-MIT License
-
-Copyright (c) 2026 Estevao
-
-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.
+MIT License
+
+Copyright (c) 2026 Estevao
+
+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.

{aru_code-0.1.0 → aru_code-0.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: aru-code
-Version: 0.1.0
+Version: 0.3.0
 Summary: A Claude Code clone built with Agno agents
 Author-email: Estevao <estevaofon@gmail.com>
 License-Expression: MIT
@@ -11,9 +11,11 @@ Keywords: ai,coding-assistant,claude,cli,agents
 Classifier: Development Status :: 3 - Alpha
 Classifier: Environment :: Console
 Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Software Development :: Libraries :: Application Frameworks
-Requires-Python: >=3.13
+Requires-Python: >=3.11
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: agno<3,>=2.5.10
@@ -50,6 +52,10 @@ An intelligent coding assistant for the terminal, powered by LLMs and [Agno](htt
 
 
 
+https://github.com/user-attachments/assets/17674bfc-3d49-4e25-bdc7-dc445d5a089d
+
+
+
 ## Highlights
 
 - **Multi-Agent Architecture** — Specialized agents for planning, execution, and conversation
@@ -65,7 +71,7 @@ An intelligent coding assistant for the terminal, powered by LLMs and [Agno](htt
 ### 1. Install
 
 ```bash
-pip install -e .
+pip install aru-code
 ```
 
 > **Requirements:** Python 3.13+
@@ -74,13 +80,7 @@ pip install -e .
 
 Aru uses **Claude Sonnet 4.6** from Anthropic as the default model. You need an [Anthropic API key](https://console.anthropic.com/) to get started.
 
-Create a `.env` file in the project root:
-
-```bash
-cp .env.example .env
-```
-
-Edit the `.env` with your key:
+Set your API key as an environment variable or create a `.env` file in your project directory:
 
 ```env
 ANTHROPIC_API_KEY=sk-ant-your-key-here
@@ -94,50 +94,7 @@ ANTHROPIC_API_KEY=sk-ant-your-key-here
 aru
 ```
 
-### Global Installation (run `aru` from anywhere)
-
-To use aru as a global command in the terminal, create a dedicated virtual environment and a wrapper script:
-
-<details>
-<summary><strong>Windows</strong></summary>
-
-1. Create the virtual environment and install:
-```bash
-python -m venv C:\aru-env
-C:\aru-env\Scripts\pip install -e C:\path\to\aru
-```
-
-2. Create `aru.bat` in a folder on your `PATH` (e.g., `C:\Users\<user>\bin\`):
-```bat
-@echo off
-C:\aru-env\Scripts\python -m aru.cli %*
-```
-
-</details>
-
-<details>
-<summary><strong>Linux / macOS</strong></summary>
-
-1. Create the virtual environment and install:
-```bash
-python3 -m venv ~/.aru-env
-~/.aru-env/bin/pip install -e /path/to/aru
-```
-
-2. Create the script `~/.local/bin/aru`:
-```bash
-#!/bin/bash
-~/.aru-env/bin/python -m aru.cli "$@"
-```
-
-3. Make it executable:
-```bash
-chmod +x ~/.local/bin/aru
-```
-
-</details>
-
-Done — now `aru` works from any directory.
+That's it — `aru` is available globally after install.
 
 ## Usage
 
@@ -187,15 +144,15 @@ By default, aru uses **Claude Sonnet 4.6** (Anthropic). You can switch to any su
 | Provider | Command | API Key (`.env`) | Extra Installation |
 |----------|---------|-------------------|------------------|
 | **Anthropic** | `/model anthropic/claude-sonnet-4-6` | `ANTHROPIC_API_KEY` | — (included) |
-| **Ollama** | `/model ollama/llama3.1` | — (local) | `pip install -e ".[ollama]"` |
-| **OpenAI** | `/model openai/gpt-4o` | `OPENAI_API_KEY` | `pip install -e ".[openai]"` |
-| **Groq** | `/model groq/llama-3.3-70b-versatile` | `GROQ_API_KEY` | `pip install -e ".[groq]"` |
-| **OpenRouter** | `/model openrouter/deepseek/deepseek-chat-v3-0324` | `OPENROUTER_API_KEY` | `pip install -e ".[openai]"` |
+| **Ollama** | `/model ollama/llama3.1` | — (local) | `pip install "aru-code[ollama]"` |
+| **OpenAI** | `/model openai/gpt-4o` | `OPENAI_API_KEY` | `pip install "aru-code[openai]"` |
+| **Groq** | `/model groq/llama-3.3-70b-versatile` | `GROQ_API_KEY` | `pip install "aru-code[groq]"` |
+| **OpenRouter** | `/model openrouter/deepseek/deepseek-chat-v3-0324` | `OPENROUTER_API_KEY` | `pip install "aru-code[openai]"` |
 
 To install all providers at once:
 
 ```bash
-pip install -e ".[all-providers]"
+pip install "aru-code[all-providers]"
 ```
 
 #### Ollama (local models)
@@ -370,7 +327,9 @@ aru-code/
 ## Development
 
 ```bash
-# Install with development dependencies
+# Clone and install in editable mode with dev dependencies
+git clone https://github.com/estevaofon/aru.git
+cd aru
 pip install -e ".[dev]"
 
 # Run tests

{aru_code-0.1.0 → aru_code-0.3.0}/README.md

@@ -6,6 +6,10 @@ An intelligent coding assistant for the terminal, powered by LLMs and [Agno](htt
 
 
 
+https://github.com/user-attachments/assets/17674bfc-3d49-4e25-bdc7-dc445d5a089d
+
+
+
 ## Highlights
 
 - **Multi-Agent Architecture** — Specialized agents for planning, execution, and conversation
@@ -21,7 +25,7 @@ An intelligent coding assistant for the terminal, powered by LLMs and [Agno](htt
 ### 1. Install
 
 ```bash
-pip install -e .
+pip install aru-code
 ```
 
 > **Requirements:** Python 3.13+
@@ -30,13 +34,7 @@ pip install -e .
 
 Aru uses **Claude Sonnet 4.6** from Anthropic as the default model. You need an [Anthropic API key](https://console.anthropic.com/) to get started.
 
-Create a `.env` file in the project root:
-
-```bash
-cp .env.example .env
-```
-
-Edit the `.env` with your key:
+Set your API key as an environment variable or create a `.env` file in your project directory:
 
 ```env
 ANTHROPIC_API_KEY=sk-ant-your-key-here
@@ -50,50 +48,7 @@ ANTHROPIC_API_KEY=sk-ant-your-key-here
 aru
 ```
 
-### Global Installation (run `aru` from anywhere)
-
-To use aru as a global command in the terminal, create a dedicated virtual environment and a wrapper script:
-
-<details>
-<summary><strong>Windows</strong></summary>
-
-1. Create the virtual environment and install:
-```bash
-python -m venv C:\aru-env
-C:\aru-env\Scripts\pip install -e C:\path\to\aru
-```
-
-2. Create `aru.bat` in a folder on your `PATH` (e.g., `C:\Users\<user>\bin\`):
-```bat
-@echo off
-C:\aru-env\Scripts\python -m aru.cli %*
-```
-
-</details>
-
-<details>
-<summary><strong>Linux / macOS</strong></summary>
-
-1. Create the virtual environment and install:
-```bash
-python3 -m venv ~/.aru-env
-~/.aru-env/bin/pip install -e /path/to/aru
-```
-
-2. Create the script `~/.local/bin/aru`:
-```bash
-#!/bin/bash
-~/.aru-env/bin/python -m aru.cli "$@"
-```
-
-3. Make it executable:
-```bash
-chmod +x ~/.local/bin/aru
-```
-
-</details>
-
-Done — now `aru` works from any directory.
+That's it — `aru` is available globally after install.
 
 ## Usage
 
@@ -143,15 +98,15 @@ By default, aru uses **Claude Sonnet 4.6** (Anthropic). You can switch to any su
 | Provider | Command | API Key (`.env`) | Extra Installation |
 |----------|---------|-------------------|------------------|
 | **Anthropic** | `/model anthropic/claude-sonnet-4-6` | `ANTHROPIC_API_KEY` | — (included) |
-| **Ollama** | `/model ollama/llama3.1` | — (local) | `pip install -e ".[ollama]"` |
-| **OpenAI** | `/model openai/gpt-4o` | `OPENAI_API_KEY` | `pip install -e ".[openai]"` |
-| **Groq** | `/model groq/llama-3.3-70b-versatile` | `GROQ_API_KEY` | `pip install -e ".[groq]"` |
-| **OpenRouter** | `/model openrouter/deepseek/deepseek-chat-v3-0324` | `OPENROUTER_API_KEY` | `pip install -e ".[openai]"` |
+| **Ollama** | `/model ollama/llama3.1` | — (local) | `pip install "aru-code[ollama]"` |
+| **OpenAI** | `/model openai/gpt-4o` | `OPENAI_API_KEY` | `pip install "aru-code[openai]"` |
+| **Groq** | `/model groq/llama-3.3-70b-versatile` | `GROQ_API_KEY` | `pip install "aru-code[groq]"` |
+| **OpenRouter** | `/model openrouter/deepseek/deepseek-chat-v3-0324` | `OPENROUTER_API_KEY` | `pip install "aru-code[openai]"` |
 
 To install all providers at once:
 
 ```bash
-pip install -e ".[all-providers]"
+pip install "aru-code[all-providers]"
 ```
 
 #### Ollama (local models)
@@ -326,7 +281,9 @@ aru-code/
 ## Development
 
 ```bash
-# Install with development dependencies
+# Clone and install in editable mode with dev dependencies
+git clone https://github.com/estevaofon/aru.git
+cd aru
 pip install -e ".[dev]"
 
 # Run tests

aru_code-0.3.0/aru/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.3.0"

{aru_code-0.1.0 → aru_code-0.3.0}/aru/agents/base.py

@@ -92,6 +92,12 @@ Define 1-10 concrete subtasks for the current step. Then execute them in order,
 calling `update_task` to mark each as "completed" or "failed" as you go. \
 When all subtasks are done, STOP. Do not add extra actions beyond the task list.
 
+## Subtask granularity — CRITICAL
+Each subtask should touch at most **3-4 files**. If the step involves many files, \
+split into subtasks grouped by concern (e.g. "Create model files", "Create route files", \
+"Update config and main"). Batch independent file writes using `write_files` or `edit_files` \
+to minimize tool calls. Batch independent file writes using `write_files` or `edit_files` to minimize tool calls.
+
 ## Guidelines
 - Read files before editing them
 - Use edit_file for targeted changes (preferred over rewriting entire files)
@@ -129,7 +135,10 @@ Only output text AFTER all subtasks are finished — a brief summary of what was
 Text output is ONLY for the final result or when you hit a blocker that needs user input.
 
 **Never retry failed shell commands with alternative syntax.** If a command fails, diagnose \
-the error — do not try `cmd /c`, absolute paths, or other wrappers hoping one works.\
+the error — do not try `cmd /c`, absolute paths, or other wrappers hoping one works.
+
+**Tool call limit**: If you see "Tool call limit reached" errors, STOP trying to use tools immediately. \
+Output a summary of what you accomplished so far and what remains. Do NOT retry rejected tool calls.\
 """
 
 # General-purpose agent (combines read + write, conversational)

aru_code-0.3.0/aru/agents/executor.py

@@ -0,0 +1,58 @@
+"""Executor agent - implements changes based on plans or direct instructions."""
+
+from agno.agent import Agent
+from agno.compression.manager import CompressionManager
+from agno.utils.log import log_warning
+
+from aru.agents.base import build_instructions
+from aru.providers import create_model
+from aru.tools.codebase import EXECUTOR_TOOLS, _get_small_model_ref
+
+# Max chars for truncation fallback when compression fails
+_TRUNCATE_FALLBACK = 3000
+
+
+class _SafeCompressionManager(CompressionManager):
+    """CompressionManager that truncates on failure instead of leaving messages uncompressed.
+
+    Agno's default behavior: if compression returns None, the message stays with
+    compressed_content=None → should_compress() fires again → infinite retry loop.
+    This subclass marks failed messages with a truncated version so the loop moves on.
+    """
+
+    async def acompress(self, messages, run_metrics=None):
+        # Track which messages are currently uncompressed
+        before = {id(m) for m in messages if m.role == "tool" and m.compressed_content is None}
+        await super().acompress(messages, run_metrics=run_metrics)
+        # Any message still uncompressed after super() = compression failed
+        for msg in messages:
+            if id(msg) in before and msg.compressed_content is None:
+                content_str = str(msg.content or "")
+                msg.compressed_content = content_str[:_TRUNCATE_FALLBACK] + (
+                    "... [truncated, compression failed]" if len(content_str) > _TRUNCATE_FALLBACK else ""
+                )
+                log_warning(f"Compression fallback (truncate) for {msg.tool_name}")
+
+
+def create_executor(model_ref: str = "anthropic/claude-sonnet-4-5", extra_instructions: str = "") -> Agent:
+    """Create and return the executor agent.
+
+    Args:
+        model_ref: Provider/model reference (e.g., "anthropic/claude-sonnet-4-5", "ollama/llama3.1").
+        extra_instructions: Additional instructions to append.
+    """
+    return Agent(
+        name="Executor",
+        model=create_model(model_ref, max_tokens=8192),
+        tools=EXECUTOR_TOOLS,
+        instructions=build_instructions("executor", extra_instructions),
+        markdown=True,
+        # Compress tool results after 5 uncompressed tool calls to save tokens
+        compress_tool_results=True,
+        compression_manager=_SafeCompressionManager(
+            model=create_model(_get_small_model_ref(), max_tokens=2048),
+            compress_tool_results=True,
+            compress_tool_results_limit=15,
+        ),
+        tool_call_limit=None,
+    )

{aru_code-0.1.0 → aru_code-0.3.0}/aru/cli.py

@@ -1,5 +1,7 @@
 """Interactive CLI for aru - a Claude Code clone."""
 
+from __future__ import annotations
+
 import asyncio
 import hashlib
 import json
@@ -982,6 +984,7 @@ def _show_help(config: AgentConfig | None):
     table.add_row("/sessions", "List recent sessions")
     table.add_row("/commands", "List custom commands")
     table.add_row("/skills", "List available skills")
+    table.add_row("/mcp", "List loaded MCP tools")
     table.add_row("/help", "Show this help")
     table.add_row("/quit", "Exit aru")
     table.add_row("! <cmd>", "Run shell command")
@@ -1068,6 +1071,7 @@ class AgentRunResult:
     """Result from run_agent_capture including text output and tool call history."""
     content: str | None = None
     tool_calls: list[str] = field(default_factory=list)
+    stalled: bool = False
 
     def with_tools_summary(self) -> str | None:
         """Return content with appended tool call summary for session history."""
@@ -1254,6 +1258,7 @@ async def run_agent_capture(agent, message: str, session: "Session | None" = Non
     console.print()
     final_content = None
     collected_tool_calls: list[str] = []
+    _stalled = False
 
     try:
         from aru.tools.codebase import set_display, set_live
@@ -1318,12 +1323,20 @@ async def run_agent_capture(agent, message: str, session: "Session | None" = Non
             tasklist_set_live(live)
             tasklist_set_display(display)
             accumulated = ""
+            # Stall detection: count consecutive events without useful work
+            # When tool_call_limit is hit, Agno silently rejects tool calls and
+            # keeps calling the model in a loop. We detect this by tracking
+            # consecutive events that aren't tool starts/completions/content.
+            _stall_counter = 0
+            _stalled = False
+            _STALL_LIMIT = 20  # break after 20 non-useful events (prevents infinite loop)
             async for event in agent.arun(agent_input, stream=True, stream_events=True, yield_run_output=True):
                 if isinstance(event, RunOutput):
                     run_output = event
                     break
 
                 if isinstance(event, ToolCallStartedEvent):
+                    _stall_counter = 0
                     if hasattr(event, "tool") and event.tool:
                         tool_name = event.tool.tool_name or "tool"
                         tool_args = event.tool.tool_args or None
@@ -1345,6 +1358,7 @@ async def run_agent_capture(agent, message: str, session: "Session | None" = Non
                     live.update(display)
 
                 elif isinstance(event, ToolCallCompletedEvent):
+                    _stall_counter = 0
                     if hasattr(event, "tool") and event.tool:
                         tool_id = getattr(event.tool, "tool_call_id", None) or getattr(event.tool, "tool_name", "tool")
                     else:
@@ -1365,6 +1379,7 @@ async def run_agent_capture(agent, message: str, session: "Session | None" = Non
                     live.update(display)
 
                 elif isinstance(event, RunContentEvent):
+                    _stall_counter = 0
                     if hasattr(event, "content") and event.content:
                         accumulated += event.content
                         unflushed = accumulated[display._flushed_len:]
@@ -1391,6 +1406,17 @@ async def run_agent_capture(agent, message: str, session: "Session | None" = Non
                         display.set_content(accumulated)
                         live.update(display)
 
+                else:
+                    # Unknown/internal event (e.g. model cycling after tool_call_limit)
+                    _stall_counter += 1
+                    if _stall_counter >= _STALL_LIMIT:
+                        _stalled = True
+                        live.console.print(
+                            "[yellow]Agent stalled (tool call limit likely reached). "
+                            "Moving on.[/yellow]"
+                        )
+                        break
+
         set_live(None)
         set_display(None)
 
@@ -1425,7 +1451,7 @@ async def run_agent_capture(agent, message: str, session: "Session | None" = Non
         console.print(f"[red]Error: {escape(str(e))}[/red]")
 
     console.print()
-    return AgentRunResult(content=final_content, tool_calls=collected_tool_calls)
+    return AgentRunResult(content=final_content, tool_calls=collected_tool_calls, stalled=_stalled)
 
 
 def ask_yes_no(prompt: str) -> bool:
@@ -1533,6 +1559,59 @@ async def execute_plan_steps(session: Session, executor_factory) -> str | None:
             run_result = await run_agent_capture(executor, step_prompt, session, lightweight=True)
             content = run_result.content
 
+            # If agent stalled (tool call limit loop), show progress and let user decide
+            if run_result.stalled:
+                from aru.tools.tasklist import get_task_store
+                store = get_task_store()
+                all_tasks = store.get_all()
+                done = [t for t in all_tasks if t["status"] == "completed"]
+                pending = [t for t in all_tasks if t["status"] not in ("completed", "failed")]
+
+                console.print(f"\n[yellow]Step {step.index} stalled (tool call limit reached).[/yellow]")
+                if done:
+                    console.print(f"  [green]Completed:[/green] {len(done)}/{len(all_tasks)} subtasks")
+                if pending:
+                    console.print(f"  [yellow]Pending:[/yellow]")
+                    for t in pending:
+                        console.print(f"    - {t.get('description', t.get('id', '?'))}")
+
+                if get_skip_permissions():
+                    step.status = "failed"
+                    continue
+
+                console.print("\n[bold]Options:[/bold]")
+                console.print("  [cyan](r)[/cyan] Retry step with additional instructions")
+                console.print("  [cyan](s)[/cyan] Skip to next step")
+                console.print("  [cyan](a)[/cyan] Abort plan execution")
+                try:
+                    choice = console.input("[bold yellow]Choice (r/s/a):[/bold yellow] ").strip().lower()
+                except (EOFError, KeyboardInterrupt):
+                    choice = "a"
+
+                if choice in ("r", "retry"):
+                    try:
+                        extra = console.input("[bold cyan]Additional instructions:[/bold cyan] ").strip()
+                    except (EOFError, KeyboardInterrupt):
+                        extra = ""
+                    if extra:
+                        # Re-run the same step with user's extra context appended
+                        step.status = "in_progress"
+                        reset_task_store()
+                        retry_prompt = step_prompt + f"\n\n## Additional Instructions\n{extra}"
+                        executor = executor_factory()
+                        run_result = await run_agent_capture(executor, retry_prompt, session, lightweight=True)
+                        content = run_result.content
+                        # Fall through to normal completion check below
+                    else:
+                        step.status = "failed"
+                        continue
+                elif choice in ("s", "skip"):
+                    step.status = "failed"
+                    continue
+                else:
+                    step.status = "failed"
+                    break
+
             # Check task store as ground truth for step completion
             from aru.tools.tasklist import get_task_store
             store = get_task_store()
@@ -1665,6 +1744,7 @@ async def run_cli(skip_permissions: bool = False, resume_id: str | None = None):
         # Apply default model from config if set
         if config.model_defaults.get("default"):
             session.model_ref = config.model_defaults["default"]
+        _sync_model(session)
         _render_home(session, skip_permissions)
 
     # Wire file-mutation callback so context cache (tree/git) is invalidated

{aru_code-0.1.0 → aru_code-0.3.0}/aru/config.py

@@ -8,6 +8,8 @@ Supports:
 Follows the Gemini .agents convention for cross-platform compatibility.
 """
 
+from __future__ import annotations
+
 import json
 import os
 from dataclasses import dataclass, field

{aru_code-0.1.0 → aru_code-0.3.0}/aru/providers.py

@@ -219,6 +219,7 @@ def resolve_model_ref(model_ref: str) -> tuple[str, str]:
 
     if "/" in model_ref:
         provider_key, model_name = model_ref.split("/", 1)
+        provider_key = provider_key.lower()
     else:
         # Could be a provider name (use its default) or unknown
         if model_ref in _providers:

{aru_code-0.1.0 → aru_code-0.3.0}/aru/tools/ast_tools.py

@@ -1,5 +1,7 @@
 """AST-based code analysis tools using tree-sitter."""
 
+from __future__ import annotations
+
 import os
 import re
 from typing import Any

{aru_code-0.1.0 → aru_code-0.3.0}/aru/tools/gitignore.py

@@ -1,5 +1,7 @@
 """Gitignore-aware file filtering for codebase operations."""
 
+from __future__ import annotations
+
 import os
 from typing import Iterator
 

{aru_code-0.1.0 → aru_code-0.3.0}/aru/tools/mcp_client.py

@@ -1,5 +1,7 @@
 """Model Context Protocol (MCP) client manager and tool generation."""
 
+from __future__ import annotations
+
 import asyncio
 import json
 import os

{aru_code-0.1.0 → aru_code-0.3.0}/aru/tools/tasklist.py

@@ -5,6 +5,8 @@ to plan and track subtasks within each plan step. Inspired by Claude Code
 and Antigravity's task management approach.
 """
 
+from __future__ import annotations
+
 import threading
 
 from rich.console import Console, Group

{aru_code-0.1.0 → aru_code-0.3.0}/aru_code.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: aru-code
-Version: 0.1.0
+Version: 0.3.0
 Summary: A Claude Code clone built with Agno agents
 Author-email: Estevao <estevaofon@gmail.com>
 License-Expression: MIT
@@ -11,9 +11,11 @@ Keywords: ai,coding-assistant,claude,cli,agents
 Classifier: Development Status :: 3 - Alpha
 Classifier: Environment :: Console
 Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Software Development :: Libraries :: Application Frameworks
-Requires-Python: >=3.13
+Requires-Python: >=3.11
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: agno<3,>=2.5.10
@@ -50,6 +52,10 @@ An intelligent coding assistant for the terminal, powered by LLMs and [Agno](htt
 
 
 
+https://github.com/user-attachments/assets/17674bfc-3d49-4e25-bdc7-dc445d5a089d
+
+
+
 ## Highlights
 
 - **Multi-Agent Architecture** — Specialized agents for planning, execution, and conversation
@@ -65,7 +71,7 @@ An intelligent coding assistant for the terminal, powered by LLMs and [Agno](htt
 ### 1. Install
 
 ```bash
-pip install -e .
+pip install aru-code
 ```
 
 > **Requirements:** Python 3.13+
@@ -74,13 +80,7 @@ pip install -e .
 
 Aru uses **Claude Sonnet 4.6** from Anthropic as the default model. You need an [Anthropic API key](https://console.anthropic.com/) to get started.
 
-Create a `.env` file in the project root:
-
-```bash
-cp .env.example .env
-```
-
-Edit the `.env` with your key:
+Set your API key as an environment variable or create a `.env` file in your project directory:
 
 ```env
 ANTHROPIC_API_KEY=sk-ant-your-key-here
@@ -94,50 +94,7 @@ ANTHROPIC_API_KEY=sk-ant-your-key-here
 aru
 ```
 
-### Global Installation (run `aru` from anywhere)
-
-To use aru as a global command in the terminal, create a dedicated virtual environment and a wrapper script:
-
-<details>
-<summary><strong>Windows</strong></summary>
-
-1. Create the virtual environment and install:
-```bash
-python -m venv C:\aru-env
-C:\aru-env\Scripts\pip install -e C:\path\to\aru
-```
-
-2. Create `aru.bat` in a folder on your `PATH` (e.g., `C:\Users\<user>\bin\`):
-```bat
-@echo off
-C:\aru-env\Scripts\python -m aru.cli %*
-```
-
-</details>
-
-<details>
-<summary><strong>Linux / macOS</strong></summary>
-
-1. Create the virtual environment and install:
-```bash
-python3 -m venv ~/.aru-env
-~/.aru-env/bin/pip install -e /path/to/aru
-```
-
-2. Create the script `~/.local/bin/aru`:
-```bash
-#!/bin/bash
-~/.aru-env/bin/python -m aru.cli "$@"
-```
-
-3. Make it executable:
-```bash
-chmod +x ~/.local/bin/aru
-```
-
-</details>
-
-Done — now `aru` works from any directory.
+That's it — `aru` is available globally after install.
 
 ## Usage
 
@@ -187,15 +144,15 @@ By default, aru uses **Claude Sonnet 4.6** (Anthropic). You can switch to any su
 | Provider | Command | API Key (`.env`) | Extra Installation |
 |----------|---------|-------------------|------------------|
 | **Anthropic** | `/model anthropic/claude-sonnet-4-6` | `ANTHROPIC_API_KEY` | — (included) |
-| **Ollama** | `/model ollama/llama3.1` | — (local) | `pip install -e ".[ollama]"` |
-| **OpenAI** | `/model openai/gpt-4o` | `OPENAI_API_KEY` | `pip install -e ".[openai]"` |
-| **Groq** | `/model groq/llama-3.3-70b-versatile` | `GROQ_API_KEY` | `pip install -e ".[groq]"` |
-| **OpenRouter** | `/model openrouter/deepseek/deepseek-chat-v3-0324` | `OPENROUTER_API_KEY` | `pip install -e ".[openai]"` |
+| **Ollama** | `/model ollama/llama3.1` | — (local) | `pip install "aru-code[ollama]"` |
+| **OpenAI** | `/model openai/gpt-4o` | `OPENAI_API_KEY` | `pip install "aru-code[openai]"` |
+| **Groq** | `/model groq/llama-3.3-70b-versatile` | `GROQ_API_KEY` | `pip install "aru-code[groq]"` |
+| **OpenRouter** | `/model openrouter/deepseek/deepseek-chat-v3-0324` | `OPENROUTER_API_KEY` | `pip install "aru-code[openai]"` |
 
 To install all providers at once:
 
 ```bash
-pip install -e ".[all-providers]"
+pip install "aru-code[all-providers]"
 ```
 
 #### Ollama (local models)
@@ -370,7 +327,9 @@ aru-code/
 ## Development
 
 ```bash
-# Install with development dependencies
+# Clone and install in editable mode with dev dependencies
+git clone https://github.com/estevaofon/aru.git
+cd aru
 pip install -e ".[dev]"
 
 # Run tests

{aru_code-0.1.0 → aru_code-0.3.0}/pyproject.toml

@@ -4,11 +4,11 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "aru-code"
-version = "0.1.0"
+version = "0.3.0"
 description = "A Claude Code clone built with Agno agents"
 readme = "README.md"
 license = "MIT"
-requires-python = ">=3.13"
+requires-python = ">=3.11"
 authors = [
     { name = "Estevao", email = "estevaofon@gmail.com" },
 ]
@@ -17,7 +17,9 @@ classifiers = [
     "Development Status :: 3 - Alpha",
     "Environment :: Console",
     "Intended Audience :: Developers",
-"Programming Language :: Python :: 3.13",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
     "Topic :: Software Development :: Libraries :: Application Frameworks",
 ]
 

aru_code-0.1.0/aru/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.1.0"

aru_code-0.1.0/aru/agents/executor.py

@@ -1,32 +0,0 @@
-"""Executor agent - implements changes based on plans or direct instructions."""
-
-from agno.agent import Agent
-from agno.compression.manager import CompressionManager
-
-from aru.agents.base import build_instructions
-from aru.providers import create_model
-from aru.tools.codebase import EXECUTOR_TOOLS, _get_small_model_ref
-
-
-def create_executor(model_ref: str = "anthropic/claude-sonnet-4-5", extra_instructions: str = "") -> Agent:
-    """Create and return the executor agent.
-
-    Args:
-        model_ref: Provider/model reference (e.g., "anthropic/claude-sonnet-4-5", "ollama/llama3.1").
-        extra_instructions: Additional instructions to append.
-    """
-    return Agent(
-        name="Executor",
-        model=create_model(model_ref, max_tokens=8192),
-        tools=EXECUTOR_TOOLS,
-        instructions=build_instructions("executor", extra_instructions),
-        markdown=True,
-        # Compress tool results after 5 uncompressed tool calls to save tokens
-        compress_tool_results=True,
-        compression_manager=CompressionManager(
-            model=create_model(_get_small_model_ref(), max_tokens=1024),
-            compress_tool_results=True,
-            compress_tool_results_limit=15,
-        ),
-        tool_call_limit=15,
-    )