Ideas to be possibly implemented in the future:
 
+** Implement new task template concept
+
+Example:
+#+begin_src yaml
+
+# AI model alias (references a configured model)
+model: default
+
+# The main prompt template (multiline string preserved with indentation)
+prompt: |
+  <|im_start|>system
+  Your task is to analyze user provided story and write the summary that
+  preserves the most valuable events, facts, conclusions or take-aways
+  from it.
+  <|im_end|>
+  <|im_start|>user
+  /think Summarize following:
+  <TASK-FILE>
+  <|im_end|>
+  <|im_start|>assistant
+#+end_src
+
 ** Documentation
 
 - Add example problem statements and resulting solutions on various
 
 ** System operation
 
-- Consider implementing a plugin architecture to allow third-party
-  developers to extend Älyverkko CLI's functionality with custom
-  modules or integrations.
-
 - Possibility to easily pause and resume Älyverkko CLI without loosing
   in-progress computation. Unix process stop and continue signals
   could possibly be used.
 
-- Explain how to monitor system performance and resource usage during
-  AI processing tasks.
-
-- Introduce a comprehensive suite of automated tests to ensure the
-  reliability and stability of new features before they are released.
-
-- Address potential security concerns, such as handling sensitive data
-  in AI processing tasks and securing communication with external
-  services.
-
 ** Data management
 
 - in maildir ignore binary files, use joinfiles command as example how
 - Implement a logging framework with configurable log levels to help
   users debug issues without cluttering the standard output.
 
-** Integration with external services
-
-- Add capabilities to connect with Jira, fetch content, and
-  potentially update issues or comments based on AI processing
-  results.
-- Implement similar integration with Confluence for content retrieval
-  and updates.
-- Extend the application's reach by adding the ability to interact
-  with arbitrary web sites, enabling information extraction and
-  automation of web-based tasks.
-
-** Tooling enhancements
-
-- Incorporate Python script execution capabilities directly by the AI,
-  expanding the range of available data manipulation and processing
-  tools.
-
-- Integrate relational database access to leverage structured data
-  storage and retrieval in workflows.
-
-- Enable web request functionality to interact with RESTful APIs or
-  scrape web content as part of task execution.
-
-- Introduce a notebook feature that allows the AI to maintain and
-  reference its own notes, fostering context retention across tasks.
-
 ** Multimedia processing
 
 - Extend the application's capabilities to include voice capture and
 
 - Implement image capture and processing features, enabling tasks that
   involve image analysis or content extraction from visual data.
-
-** Task preparation and queue management
-
-- Refactor the task queue mechanism to support:
-  - Multiple task sources, including a REST API endpoint for
-    programmatic task submission.
-  - Load balancing across multiple executors (instances of Älyverkko
-    CLI) with dynamic registration and unregistration without system
-    interruption.
-  - Task priority assignments, ensuring that critical tasks are
-    processed in a timely manner.
-
-- Offer guidance on preparing input files for batch processing,
-  including best practices for formatting data and structuring problem
-  statements.
-
-** User interface development
-
-- Create a web-based UI to provide users with an interface for task
-  submission and result retrieval, improving accessibility and user
-  experience.
-
-- Integrate Quality of Service (QoS) concepts within the UI to ensure
-  equitable resource allocation among users.
-
-- Implement administrative features for managing user accounts and
-  system resources, maintaining a secure and efficient operating
-  environment.