gac 2.7.3__tar.gz → 3.12.0__tar.gz

{gac-2.7.3 → gac-3.12.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gac
-Version: 2.7.3
+Version: 3.12.0
 Summary: LLM-powered Git commit message generator with multi-provider support
 Project-URL: Homepage, https://github.com/cellwebb/gac
 Project-URL: Documentation, https://github.com/cellwebb/gac#readme
@@ -22,7 +22,6 @@ Classifier: Programming Language :: Python :: Implementation :: CPython
 Classifier: Programming Language :: Python :: Implementation :: PyPy
 Requires-Python: >=3.10
 Requires-Dist: click>=8.3.0
-Requires-Dist: halo
 Requires-Dist: httpcore>=1.0.9
 Requires-Dist: httpx>=0.28.0
 Requires-Dist: prompt-toolkit>=3.0.36
@@ -30,7 +29,6 @@ Requires-Dist: pydantic>=2.12.0
 Requires-Dist: python-dotenv>=1.1.1
 Requires-Dist: questionary
 Requires-Dist: rich>=14.1.0
-Requires-Dist: tiktoken>=0.12.0
 Provides-Extra: dev
 Requires-Dist: build; extra == 'dev'
 Requires-Dist: codecov; extra == 'dev'
@@ -48,7 +46,7 @@ Description-Content-Type: text/markdown
 # 🚀 Git Auto Commit (gac)
 
 [![PyPI version](https://img.shields.io/pypi/v/gac.svg)](https://pypi.org/project/gac/)
-[![Python](https://img.shields.io/badge/python-3.10%20|%203.11%20|%203.12%20|%203.13%20|%203.14-blue.svg)](https://www.python.org/downloads/)
+[![Python](https://img.shields.io/badge/python-3.10--3.14-blue.svg)](https://www.python.org/downloads/)
 [![Build Status](https://github.com/cellwebb/gac/actions/workflows/ci.yml/badge.svg)](https://github.com/cellwebb/gac/actions)
 [![codecov](https://codecov.io/gh/cellwebb/gac/branch/main/graph/badge.svg)](https://app.codecov.io/gh/cellwebb/gac)
 [![Ruff](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/ruff/main/assets/badge/v2.json)](https://github.com/astral-sh/ruff)
@@ -105,11 +103,12 @@ uv tool upgrade gac
 
 ## Key Features
 
-### 🌐 **Supported Providers**
+### 🌐 **25+ Supported Providers**
 
-- **Anthropic** • **Cerebras** • **Chutes.ai** • **Claude Code** • **DeepSeek**
-- **Fireworks** • **Gemini** • **Groq** • **LM Studio** • **MiniMax** • **Mistral** • **Ollama**
-- **OpenAI** • **OpenRouter** • **Streamlake** • **Synthetic.new** • **Together AI**
+- **Anthropic** • **Azure OpenAI** • **Cerebras** • **Chutes.ai** • **Claude Code (OAuth)**
+- **DeepSeek** • **Fireworks** • **Gemini** • **Groq** • **Kimi for Coding** • **LM Studio**
+- **MiniMax.io** • **Mistral AI** • **Moonshot AI** • **Ollama** • **OpenAI** • **OpenRouter**
+- **Qwen.ai (OAuth)** • **Replicate** • **Streamlake** • **Synthetic.new** • **Together AI**
 - **Z.AI** • **Z.AI Coding** • **Custom Endpoints (Anthropic/OpenAI)**
 
 ### 🧠 **Smart LLM Analysis**
@@ -130,11 +129,12 @@ uv tool upgrade gac
 - **25+ languages**: Generate commit messages in English, Chinese, Japanese, Korean, Spanish, French, German, and 20+ more languages
 - **Flexible translation**: Choose to keep conventional commit prefixes in English for tool compatibility, or fully translate them
 - **Multiple workflows**: Set a default language with `gac language`, or use `-l <language>` flag for one-time overrides
-- **Native script support**: Full support for non-Latin scripts including CJK, Cyrillic, Arabic, and more
+- **Native script support**: Full support for non-Latin scripts including CJK, Cyrillic, Thai, and more
 
 ### 💻 **Developer Experience**
 
 - **Interactive feedback**: Type `r` to reroll, `e` to edit in-place with vi/emacs keybindings, or directly type your feedback like `make it shorter` or `focus on the bug fix`
+- **Interactive questioning**: Use `--interactive` (`-i`) to answer targeted questions about your changes for more contextual commit messages
 - **One-command workflows**: Complete workflows with flags like `gac -ayp` (stage all, auto-confirm, push)
 - **Git integration**: Respects pre-commit and lefthook hooks, running them before expensive LLM operations
 
@@ -171,6 +171,7 @@ gac
 | `gac -v`        | Verbose format with Motivation, Technical Approach, and Impact Analysis |
 | `gac -h "hint"` | Add context for LLM (e.g., `gac -h "bug fix"`)                          |
 | `gac -s`        | Include scope (e.g., feat(auth):)                                       |
+| `gac -i`        | Ask questions about changes for better context                          |
 | `gac -p`        | Commit and push                                                         |
 
 ### Power User Examples
@@ -188,6 +189,9 @@ gac -o
 # Group changes into logically related commits
 gac -ag
 
+# Interactive mode with verbose output for detailed explanations
+gac -iv
+
 # Debug what the LLM sees
 gac --show-prompt
 
@@ -257,6 +261,8 @@ Track real-time installation metrics and package download statistics.
 ## Getting Help
 
 - **Full documentation**: [docs/USAGE.md](docs/en/USAGE.md) - Complete CLI reference
+- **Claude Code OAuth**: [docs/CLAUDE_CODE.md](docs/en/CLAUDE_CODE.md) - Claude Code setup and authentication
+- **Qwen.ai OAuth**: [docs/QWEN.md](docs/en/QWEN.md) - Qwen.ai setup and authentication
 - **Custom prompts**: [docs/CUSTOM_SYSTEM_PROMPTS.md](docs/en/CUSTOM_SYSTEM_PROMPTS.md) - Customize commit message style
 - **Troubleshooting**: [docs/TROUBLESHOOTING.md](docs/en/TROUBLESHOOTING.md) - Common issues and solutions
 - **Contributing**: [docs/CONTRIBUTING.md](docs/en/CONTRIBUTING.md) - Development setup and guidelines

{gac-2.7.3 → gac-3.12.0}/README.md

@@ -6,7 +6,7 @@
 # 🚀 Git Auto Commit (gac)
 
 [![PyPI version](https://img.shields.io/pypi/v/gac.svg)](https://pypi.org/project/gac/)
-[![Python](https://img.shields.io/badge/python-3.10%20|%203.11%20|%203.12%20|%203.13%20|%203.14-blue.svg)](https://www.python.org/downloads/)
+[![Python](https://img.shields.io/badge/python-3.10--3.14-blue.svg)](https://www.python.org/downloads/)
 [![Build Status](https://github.com/cellwebb/gac/actions/workflows/ci.yml/badge.svg)](https://github.com/cellwebb/gac/actions)
 [![codecov](https://codecov.io/gh/cellwebb/gac/branch/main/graph/badge.svg)](https://app.codecov.io/gh/cellwebb/gac)
 [![Ruff](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/ruff/main/assets/badge/v2.json)](https://github.com/astral-sh/ruff)
@@ -63,11 +63,12 @@ uv tool upgrade gac
 
 ## Key Features
 
-### 🌐 **Supported Providers**
+### 🌐 **25+ Supported Providers**
 
-- **Anthropic** • **Cerebras** • **Chutes.ai** • **Claude Code** • **DeepSeek**
-- **Fireworks** • **Gemini** • **Groq** • **LM Studio** • **MiniMax** • **Mistral** • **Ollama**
-- **OpenAI** • **OpenRouter** • **Streamlake** • **Synthetic.new** • **Together AI**
+- **Anthropic** • **Azure OpenAI** • **Cerebras** • **Chutes.ai** • **Claude Code (OAuth)**
+- **DeepSeek** • **Fireworks** • **Gemini** • **Groq** • **Kimi for Coding** • **LM Studio**
+- **MiniMax.io** • **Mistral AI** • **Moonshot AI** • **Ollama** • **OpenAI** • **OpenRouter**
+- **Qwen.ai (OAuth)** • **Replicate** • **Streamlake** • **Synthetic.new** • **Together AI**
 - **Z.AI** • **Z.AI Coding** • **Custom Endpoints (Anthropic/OpenAI)**
 
 ### 🧠 **Smart LLM Analysis**
@@ -88,11 +89,12 @@ uv tool upgrade gac
 - **25+ languages**: Generate commit messages in English, Chinese, Japanese, Korean, Spanish, French, German, and 20+ more languages
 - **Flexible translation**: Choose to keep conventional commit prefixes in English for tool compatibility, or fully translate them
 - **Multiple workflows**: Set a default language with `gac language`, or use `-l <language>` flag for one-time overrides
-- **Native script support**: Full support for non-Latin scripts including CJK, Cyrillic, Arabic, and more
+- **Native script support**: Full support for non-Latin scripts including CJK, Cyrillic, Thai, and more
 
 ### 💻 **Developer Experience**
 
 - **Interactive feedback**: Type `r` to reroll, `e` to edit in-place with vi/emacs keybindings, or directly type your feedback like `make it shorter` or `focus on the bug fix`
+- **Interactive questioning**: Use `--interactive` (`-i`) to answer targeted questions about your changes for more contextual commit messages
 - **One-command workflows**: Complete workflows with flags like `gac -ayp` (stage all, auto-confirm, push)
 - **Git integration**: Respects pre-commit and lefthook hooks, running them before expensive LLM operations
 
@@ -129,6 +131,7 @@ gac
 | `gac -v`        | Verbose format with Motivation, Technical Approach, and Impact Analysis |
 | `gac -h "hint"` | Add context for LLM (e.g., `gac -h "bug fix"`)                          |
 | `gac -s`        | Include scope (e.g., feat(auth):)                                       |
+| `gac -i`        | Ask questions about changes for better context                          |
 | `gac -p`        | Commit and push                                                         |
 
 ### Power User Examples
@@ -146,6 +149,9 @@ gac -o
 # Group changes into logically related commits
 gac -ag
 
+# Interactive mode with verbose output for detailed explanations
+gac -iv
+
 # Debug what the LLM sees
 gac --show-prompt
 
@@ -215,6 +221,8 @@ Track real-time installation metrics and package download statistics.
 ## Getting Help
 
 - **Full documentation**: [docs/USAGE.md](docs/en/USAGE.md) - Complete CLI reference
+- **Claude Code OAuth**: [docs/CLAUDE_CODE.md](docs/en/CLAUDE_CODE.md) - Claude Code setup and authentication
+- **Qwen.ai OAuth**: [docs/QWEN.md](docs/en/QWEN.md) - Qwen.ai setup and authentication
 - **Custom prompts**: [docs/CUSTOM_SYSTEM_PROMPTS.md](docs/en/CUSTOM_SYSTEM_PROMPTS.md) - Customize commit message style
 - **Troubleshooting**: [docs/TROUBLESHOOTING.md](docs/en/TROUBLESHOOTING.md) - Common issues and solutions
 - **Contributing**: [docs/CONTRIBUTING.md](docs/en/CONTRIBUTING.md) - Development setup and guidelines

{gac-2.7.3 → gac-3.12.0}/pyproject.toml

@@ -29,8 +29,7 @@ dependencies = [
     "httpx>=0.28.0",
     "httpcore>=1.0.9",  # Required for Python 3.14 compatibility
 
-    # Token counting (OpenAI models)
-    "tiktoken>=0.12.0",
+
 
     # Core functionality
     "pydantic>=2.12.0",
@@ -38,7 +37,6 @@ dependencies = [
 
     # CLI and formatting
     "click>=8.3.0",
-    "halo",
     "questionary",
     "rich>=14.1.0",
     "prompt_toolkit>=3.0.36",
@@ -203,8 +201,8 @@ addopts = "-m 'not integration'"
 python_version = "3.10"
 warn_return_any = true
 warn_unused_configs = true
-disallow_untyped_defs = false
-disallow_incomplete_defs = false
+disallow_untyped_defs = true
+disallow_incomplete_defs = true
 check_untyped_defs = true
 no_implicit_optional = true
 warn_redundant_casts = true
@@ -218,10 +216,6 @@ show_error_codes = true
 module = "gac.providers.*"
 warn_return_any = false
 
-[[tool.mypy.overrides]]
-module = "halo"
-ignore_missing_imports = true
-
 [template.plugins.default]
 tests = true
 src-layout = true

{gac-2.7.3 → gac-3.12.0}/src/gac/__init__.py

@@ -1,15 +1,13 @@
 """Git Auto Commit (gac) - Generate commit messages using AI."""
 
+from gac import init_cli
 from gac.__version__ import __version__
 from gac.ai import generate_commit_message
-from gac.git import get_staged_files, push_changes
-from gac.prompt import build_prompt, clean_commit_message
+from gac.prompt import build_prompt
 
 __all__ = [
     "__version__",
-    "generate_commit_message",
     "build_prompt",
-    "clean_commit_message",
-    "get_staged_files",
-    "push_changes",
+    "generate_commit_message",
+    "init_cli",
 ]

{gac-2.7.3 → gac-3.12.0}/src/gac/__version__.py

@@ -1,3 +1,3 @@
 """Version information for gac package."""
 
-__version__ = "2.7.3"
+__version__ = "3.12.0"

{gac-2.7.3 → gac-3.12.0}/src/gac/ai.py

@@ -9,29 +9,7 @@ import logging
 from gac.ai_utils import generate_with_retries
 from gac.constants import EnvDefaults
 from gac.errors import AIError
-from gac.providers import (
-    call_anthropic_api,
-    call_cerebras_api,
-    call_chutes_api,
-    call_claude_code_api,
-    call_custom_anthropic_api,
-    call_custom_openai_api,
-    call_deepseek_api,
-    call_fireworks_api,
-    call_gemini_api,
-    call_groq_api,
-    call_lmstudio_api,
-    call_minimax_api,
-    call_mistral_api,
-    call_ollama_api,
-    call_openai_api,
-    call_openrouter_api,
-    call_streamlake_api,
-    call_synthetic_api,
-    call_together_api,
-    call_zai_api,
-    call_zai_coding_api,
-)
+from gac.providers import PROVIDER_REGISTRY
 
 logger = logging.getLogger(__name__)
 
@@ -45,6 +23,7 @@ def generate_commit_message(
     quiet: bool = False,
     is_group: bool = False,
     skip_success_message: bool = False,
+    task_description: str = "commit message",
 ) -> str:
     """Generate a commit message using direct API calls to AI providers.
 
@@ -85,35 +64,10 @@ def generate_commit_message(
             {"role": "user", "content": user_prompt},
         ]
 
-    # Provider functions mapping
-    provider_funcs = {
-        "anthropic": call_anthropic_api,
-        "cerebras": call_cerebras_api,
-        "claude-code": call_claude_code_api,
-        "chutes": call_chutes_api,
-        "custom-anthropic": call_custom_anthropic_api,
-        "custom-openai": call_custom_openai_api,
-        "deepseek": call_deepseek_api,
-        "fireworks": call_fireworks_api,
-        "gemini": call_gemini_api,
-        "groq": call_groq_api,
-        "lm-studio": call_lmstudio_api,
-        "minimax": call_minimax_api,
-        "mistral": call_mistral_api,
-        "ollama": call_ollama_api,
-        "openai": call_openai_api,
-        "openrouter": call_openrouter_api,
-        "streamlake": call_streamlake_api,
-        "synthetic": call_synthetic_api,
-        "together": call_together_api,
-        "zai": call_zai_api,
-        "zai-coding": call_zai_coding_api,
-    }
-
     # Generate the commit message using centralized retry logic
     try:
         return generate_with_retries(
-            provider_funcs=provider_funcs,
+            provider_funcs=PROVIDER_REGISTRY,
             model=model,
             messages=messages,
             temperature=temperature,
@@ -122,6 +76,7 @@ def generate_commit_message(
             quiet=quiet,
             is_group=is_group,
             skip_success_message=skip_success_message,
+            task_description=task_description,
         )
     except AIError:
         # Re-raise AIError exceptions as-is to preserve error classification
@@ -150,4 +105,5 @@ def generate_grouped_commits(
         quiet=quiet,
         is_group=True,
         skip_success_message=skip_success_message,
+        task_description="commit message",
     )

{gac-2.7.3 → gac-3.12.0}/src/gac/ai_utils.py

@@ -3,44 +3,35 @@
 This module provides utility functions that support the AI provider implementations.
 """
 
+import json
 import logging
 import os
 import time
-from functools import lru_cache
-from typing import Any
+from collections.abc import Callable
+from typing import Any, cast
 
-import tiktoken
-from halo import Halo
+from rich.console import Console
+from rich.status import Status
 
-from gac.constants import EnvDefaults, Utility
 from gac.errors import AIError
+from gac.oauth import QwenOAuthProvider, refresh_token_if_expired
+from gac.oauth.token_store import TokenStore
+from gac.providers import SUPPORTED_PROVIDERS
 
 logger = logging.getLogger(__name__)
-
-
-@lru_cache(maxsize=1)
-def _should_skip_tiktoken_counting() -> bool:
-    """Return True when token counting should avoid tiktoken calls entirely."""
-    value = os.getenv("GAC_NO_TIKTOKEN", str(EnvDefaults.NO_TIKTOKEN))
-    return value.lower() in ("true", "1", "yes", "on")
+console = Console()
 
 
 def count_tokens(content: str | list[dict[str, str]] | dict[str, Any], model: str) -> int:
-    """Count tokens in content using the model's tokenizer."""
+    """Count tokens in content using character-based estimation (1 token per 3.4 characters)."""
     text = extract_text_content(content)
     if not text:
         return 0
 
-    if _should_skip_tiktoken_counting():
-        return len(text) // 4
-
-    try:
-        encoding = get_encoding(model)
-        return len(encoding.encode(text))
-    except Exception as e:
-        logger.error(f"Error counting tokens: {e}")
-        # Fallback to rough estimation (4 chars per token on average)
-        return len(text) // 4
+    # Use simple character-based estimation: 1 token per 3.4 characters (rounded)
+    result = round(len(text) / 3.4)
+    # Ensure at least 1 token for non-empty text
+    return result if result > 0 else 1
 
 
 def extract_text_content(content: str | list[dict[str, str]] | dict[str, Any]) -> str:
@@ -50,53 +41,12 @@ def extract_text_content(content: str | list[dict[str, str]] | dict[str, Any]) -
     elif isinstance(content, list):
         return "\n".join(msg["content"] for msg in content if isinstance(msg, dict) and "content" in msg)
     elif isinstance(content, dict) and "content" in content:
-        return content["content"]  # type: ignore[no-any-return]
+        return cast(str, content["content"])
     return ""
 
 
-@lru_cache(maxsize=1)
-def get_encoding(model: str) -> tiktoken.Encoding:
-    """Get the appropriate encoding for a given model."""
-    provider, model_name = model.split(":", 1) if ":" in model else (None, model)
-
-    if provider != "openai":
-        return tiktoken.get_encoding(Utility.DEFAULT_ENCODING)
-
-    try:
-        return tiktoken.encoding_for_model(model_name)
-    except KeyError:
-        # Fall back to default encoding if model not found
-        return tiktoken.get_encoding(Utility.DEFAULT_ENCODING)
-    except Exception:
-        # If there are any network/SSL issues, fall back to default encoding
-        return tiktoken.get_encoding(Utility.DEFAULT_ENCODING)
-
-
-def _classify_error(error_str: str) -> str:
-    """Classify error types based on error message content."""
-    error_str = error_str.lower()
-
-    if (
-        "api key" in error_str
-        or "unauthorized" in error_str
-        or "authentication" in error_str
-        or "invalid api key" in error_str
-    ):
-        return "authentication"
-    elif "timeout" in error_str or "timed out" in error_str or "request timeout" in error_str:
-        return "timeout"
-    elif "rate limit" in error_str or "too many requests" in error_str or "rate limit exceeded" in error_str:
-        return "rate_limit"
-    elif "connect" in error_str or "network" in error_str or "network connection failed" in error_str:
-        return "connection"
-    elif "model" in error_str or "not found" in error_str or "model not found" in error_str:
-        return "model"
-    else:
-        return "unknown"
-
-
 def generate_with_retries(
-    provider_funcs: dict,
+    provider_funcs: dict[str, Callable[..., str]],
     model: str,
     messages: list[dict[str, str]],
     temperature: float,
@@ -105,6 +55,7 @@ def generate_with_retries(
     quiet: bool = False,
     is_group: bool = False,
     skip_success_message: bool = False,
+    task_description: str = "commit message",
 ) -> str:
     """Generate content with retry logic using direct API calls."""
     # Parse model string to determine provider and actual model
@@ -114,51 +65,74 @@ def generate_with_retries(
     provider, model_name = model.split(":", 1)
 
     # Validate provider
-    supported_providers = [
-        "anthropic",
-        "cerebras",
-        "chutes",
-        "claude-code",
-        "deepseek",
-        "fireworks",
-        "gemini",
-        "groq",
-        "lm-studio",
-        "minimax",
-        "mistral",
-        "ollama",
-        "openai",
-        "openrouter",
-        "streamlake",
-        "synthetic",
-        "together",
-        "zai",
-        "zai-coding",
-        "custom-anthropic",
-        "custom-openai",
-    ]
-    if provider not in supported_providers:
-        raise AIError.model_error(f"Unsupported provider: {provider}. Supported providers: {supported_providers}")
+    if provider not in SUPPORTED_PROVIDERS:
+        raise AIError.model_error(f"Unsupported provider: {provider}. Supported providers: {SUPPORTED_PROVIDERS}")
 
     if not messages:
         raise AIError.model_error("No messages provided for AI generation")
 
+    # Load Claude Code token from TokenStore if needed
+    if provider == "claude-code":
+        # Check token expiry and refresh if needed
+        if not refresh_token_if_expired(quiet=True):
+            raise AIError.authentication_error(
+                "Claude Code token not found or expired. Please authenticate with 'gac auth claude-code login'."
+            )
+
+        # Load the (possibly refreshed) token
+        token_store = TokenStore()
+        token_data = token_store.get_token("claude-code")
+        if token_data and "access_token" in token_data:
+            os.environ["CLAUDE_CODE_ACCESS_TOKEN"] = token_data["access_token"]
+        else:
+            raise AIError.authentication_error(
+                "Claude Code token not found. Please authenticate with 'gac auth claude-code login'."
+            )
+
+    # Check Qwen OAuth token expiry and refresh if needed
+    if provider == "qwen":
+        oauth_provider = QwenOAuthProvider(TokenStore())
+        token = oauth_provider.get_token()
+        if not token:
+            if not quiet:
+                console.print("[yellow]⚠ Qwen authentication not found or expired[/yellow]")
+                console.print("[cyan]🔐 Starting automatic authentication...[/cyan]")
+            try:
+                oauth_provider.initiate_auth(open_browser=True)
+                token = oauth_provider.get_token()
+                if not token:
+                    raise AIError.authentication_error(
+                        "Qwen authentication failed. Run 'gac auth qwen login' to authenticate manually."
+                    )
+                if not quiet:
+                    console.print("[green]✓ Authentication successful![/green]\n")
+            except AIError:
+                raise
+            except (ValueError, KeyError, json.JSONDecodeError, ConnectionError, OSError) as e:
+                raise AIError.authentication_error(
+                    f"Qwen authentication failed: {e}. Run 'gac auth qwen login' to authenticate manually."
+                ) from e
+
     # Set up spinner
-    message_type = "commit messages" if is_group else "commit message"
+    if is_group:
+        message_type = f"grouped {task_description}s"
+    else:
+        message_type = task_description
+
     if quiet:
         spinner = None
     else:
-        spinner = Halo(text=f"Generating {message_type} with {provider} {model_name}...", spinner="dots")
+        spinner = Status(f"Generating {message_type} with {provider} {model_name}...")
         spinner.start()
 
-    last_exception = None
+    last_exception: Exception | None = None
     last_error_type = "unknown"
 
     for attempt in range(max_retries):
         try:
             if not quiet and not skip_success_message and attempt > 0:
                 if spinner:
-                    spinner.text = f"Retry {attempt + 1}/{max_retries} with {provider} {model_name}..."
+                    spinner.update(f"Retry {attempt + 1}/{max_retries} with {provider} {model_name}...")
                 logger.info(f"Retry attempt {attempt + 1}/{max_retries}")
 
             # Call the appropriate provider function
@@ -172,54 +146,50 @@ def generate_with_retries(
                 if skip_success_message:
                     spinner.stop()  # Stop spinner without showing success/failure
                 else:
-                    spinner.succeed(f"Generated {message_type} with {provider} {model_name}")
+                    spinner.stop()
+                    console.print(f"✓ Generated {message_type} with {provider} {model_name}")
 
             if content is not None and content.strip():
-                return content.strip()  # type: ignore[no-any-return]
+                return content.strip()
             else:
                 logger.warning(f"Empty or None content received from {provider} {model_name}: {repr(content)}")
                 raise AIError.model_error("Empty response from AI model")
 
-        except Exception as e:
+        except AIError as e:
             last_exception = e
-            error_type = _classify_error(str(e))
+            error_type = e.error_type
             last_error_type = error_type
 
             # For authentication and model errors, don't retry
             if error_type in ["authentication", "model"]:
                 if spinner and not skip_success_message:
-                    spinner.fail(f"Failed to generate {message_type} with {provider} {model_name}")
-
-                # Create the appropriate error type based on classification
-                if error_type == "authentication":
-                    raise AIError.authentication_error(f"AI generation failed: {str(e)}") from e
-                elif error_type == "model":
-                    raise AIError.model_error(f"AI generation failed: {str(e)}") from e
+                    spinner.stop()
+                    console.print(f"✗ Failed to generate {message_type} with {provider} {model_name}")
+                raise
 
             if attempt < max_retries - 1:
                 # Exponential backoff
                 wait_time = 2**attempt
                 if not quiet and not skip_success_message:
                     if attempt == 0:
-                        logger.warning(f"AI generation failed, retrying in {wait_time}s: {str(e)}")
+                        logger.warning(f"AI generation failed, retrying in {wait_time}s: {e}")
                     else:
-                        logger.warning(
-                            f"AI generation failed (attempt {attempt + 1}), retrying in {wait_time}s: {str(e)}"
-                        )
+                        logger.warning(f"AI generation failed (attempt {attempt + 1}), retrying in {wait_time}s: {e}")
 
                 if spinner and not skip_success_message:
                     for i in range(wait_time, 0, -1):
-                        spinner.text = f"Retry {attempt + 1}/{max_retries} in {i}s..."
+                        spinner.update(f"Retry {attempt + 1}/{max_retries} in {i}s...")
                         time.sleep(1)
                 else:
                     time.sleep(wait_time)
             else:
                 num_retries = max_retries
                 retry_word = "retry" if num_retries == 1 else "retries"
-                logger.error(f"AI generation failed after {num_retries} {retry_word}: {str(e)}")
+                logger.error(f"AI generation failed after {num_retries} {retry_word}: {e}")
 
     if spinner and not skip_success_message:
-        spinner.fail(f"Failed to generate {message_type} with {provider} {model_name}")
+        spinner.stop()
+        console.print(f"✗ Failed to generate {message_type} with {provider} {model_name}")
 
     # If we get here, all retries failed - use the last classified error type
     num_retries = max_retries