- Add local import of json inside raw_stream_generate() nested function
- This fixes NameError that occurred when streaming responses

- Removed incorrect 'await' on generate_stream which returns an async generator

- Added pattern for <tool_call><tool><name>...</name><parameters><key>value</key></parameters></tool></tool_call>
- Added pattern for standalone <tool><name>...</name><parameters><key>value</key></parameters></tool>
- Extracts key-value pairs from nested XML tags inside <parameters>

- Simplified ApexBig50Parser to only keep unique patterns (markdown JSON, react)
- Complex XML parsing is now handled by ToolCallParser as the fallback
- This eliminates code duplication between the two parsers

- Added pattern for <tool_call><tool><function>...</function><parameters>...</parameters></tool></tool_call>
- Added pattern for standalone <tool><function>...</function><parameters>...</parameters></tool>
- Added same patterns to ApexBig50Parser

- Added support for <tool_call><tool><name>...</name><arguments>...</arguments></tool></tool_call> format
- Added support for multiple tool calls in <tool_call> wrapper
- Added same patterns to ApexBig50Parser for direct handling
- Added exception handling for formatter.format_full() to prevent TypeError
- Added fallback response when formatted_response is None
- ToolCallParser now serves as fallback for all specific model parsers

- Added new pattern to parse <tool><action>...</action><parameters>...</parameters></tool>
- Handles both JSON and XML-style key-value pairs in parameters
- Example:
  <tool>
  <action>search</action>
  <parameters>
    <query>Apple AAPL Q4 financials Net Income Revenue</query>
  </parameters>
  </tool>

- Added new pattern to parse <tool>name</tool><tool_call>JSON</tool_call> format
- Handles both closing tag (</tool_call>) and end-of-string as terminator
- Works with formats like:
  <tool>search</tool>
  <tool_call>
  {"query": "Apple AAPL Q4 financial report"}
  </tool_call>

- Added debug output to ToolCallParser.extract_tool_calls() method
- Shows when parser is called, input text (first 200 chars), and output
- Debug logging follows existing codebase pattern using print() with DEBUG prefix
- Useful for troubleshooting tool call extraction issues

- Added _parse_xml_style_tool_calls() method to parse XML format like:
  <tool><name>search</name><arguments>{...}</arguments></tool>
- Modified extract_tool_calls() to use XML parsing as fallback
- ToolCallParser now serves as fallback for all specific model parsers
  via ModelParserAdapter (QwenParser, DeepSeekParser, LlamaParser,
  MistralParser, ClaudeParser, CommandRParser, GemmaParser, GrokParser,
  PhiParser, ApexBig50Parser)

Support for:
<tool_call>
<tool>
<name>search</name>
<arguments>{"query": "..."}</arguments>
</tool>
</tool_call>

Support for:
<tool_call>
<tool>
<name>financial_data_lookup</name>
<param>...</param>
</tool>
</tool_call>

Support for format:
<tool>yahoo_finance_api</tool>[{"symbol": "AAPL", ...}]

Support for:
<tool_call>
<tool>
<action>search</action>
<parameters>
  <query>...</query>
</parameters>
</tool>
</tool_call>

current_manager is the individual model manager which doesn't have 'models'.
multi_model_manager has the 'models' dict with all loaded models.

- Add safety check for None formatted_response in debug output
- Add debug logging when litellm fallback fails

When resolving 'default', skip keys like 'default', 'image', 'audio',
'image:*', 'audio:*' to find the first actual LLM model.

When resolving 'default' model, skip the 'default' key in models dict
and return the first actual loaded model instead.

- Removed --force-all CLI flag
- Updated help text to document 'all' option
- Updated reasoning_choices to expand 'all' to all options
- Updated parsing logic to handle 'all' as string

When requested_model is 'default', get_resolved_model_name now:
1. First tries current_manager.default_model
2. Falls back to the first model in manager.models

This ensures the model family is correctly detected for parser selection.

When the model-specific parser (like QwenParser) doesn't find any
tool calls, fall back to ToolCallParser as a catch-all.

The regex pattern was missing <tool> - it only matched <tool_call>,
funktion>, and <tool_use>. Now it also matches <tool>.

Add custom XML tool format parser for <tool><action>...</action><object>...</object><properties>...</properties></tool>

The model was generating tool calls in this format:
<tool>
<action>search</action>
<object>financial_data</object>
<properties>
  <query>...</query>
</properties>
</tool>

Added parser support in ApexBig50Parser and strip_tool_calls_from_content
to handle this custom format.

- Bug 1: Skip format_tools_for_prompt in raw mode (already had condition)
- Bug 2: Use final_text (after reasoning) instead of generated_text for formatter
- Bug 3: Pass final_text to ModelParserAdapter instead of generated_text

This prevents reasoning from appearing in both content AND reasoning fields,
and allows the tool parser to properly extract tool calls without being
confused by reasoning tags.

- Add new --force-all CLI argument
- Update --force-reasoning help text to mention --force-all
- Handle --force-all in main function to expand to all reasoning options

- Add force_reasoning_prompt function with Big 10 family prefixes
- Add inject_system and force_reasoning parameters
- Update --force-reasoning CLI with comma-separated options
- Add --dump option to show raw output, parsed output, and litellm debug
- Fix stop tokens to include ]]> when prompt is selected
- Add mock strategy for fake reasoning stats
- Chain --system-prompt at start of existing system message
- Add 'raw' option to --force-reasoning
- Fix format_tools_for_prompt to skip in raw mode
- Pass tools to format_for_raw_completion in raw mode
- Add parse_and_format method to OpenAIFormatter for tool extraction
- Use parse_and_format in raw mode for correct tool extraction pipeline

Pipeline: Model output -> Extract reasoning (raw mode) -> ModelParserAdapter (extract tools) -> OpenAIFormatter (final format)

- Add truthy check before 'in' operator to prevent TypeError when
  force_reasoning_args is None (when --force-reasoning is not specified)
- Fixes: name 'force_reasoning_args' is not defined error

- In raw mode, extracted reasoning is now preserved in the response
- Mock reasoning is only applied when there's no existing reasoning
- Added logic to set extracted reasoning in message after formatter
- Same fix applied to non-raw path in generate_chat_response

- System prompt now includes: 'CRITICAL: You must always close your reasoning with ]]> before opening any tool tags'
- Extraction logic now uses tool tags as fallback stop markers if close tag is missing
- Handles: <tool_call>, <tool>, <|tool_call|>, <|tool|>, <function=

Shows:
- Full first pass result
- Extraction details (close tag used, reasoning text, final text)
- Cleanup details

Now raw mode passes the generated text through OpenAIFormatter which:
- Handles tool extraction
- Provides OpenAI compatibility
- Handles other post-processing

This ensures raw mode output is treated the same as regular mode.

- Add cleanup_control_tokens function to strip leading/trailing control tokens
- Apply cleanup to final_text and second_pass_result in raw mode
- Add mock strategy handling to raw mode (was missing)
- Add debug output for cleanup steps

When 'raw' is used without 'prompt', template_manager wasn't defined.
Now creating it on-demand when needed.

When 'raw' is used, skip the 'prompt', 'inject', and 'stop' handlers
since raw mode handles everything separately. This was causing
double assistant headers and corrupted prompts.

The tokenizer approach was causing double assistant headers.
Now using only template_manager.format_for_raw_completion which
handles everything correctly.

The AgenticTemplateManager already has a format_for_raw_completion method
that handles prompt formatting with reasoning tags. No need to manually
find the tokenizer - just use the existing template logic.

Now shows:
- current_manager type and backend type
- Available attributes on the backend
- Which path was used to find (or not find) the tokenizer
- Also checks model_manager.tokenizer as fallback