Dev #117
Merged
Dev #117
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Added: - Basic Cog implementation - A few basic unittest for the cogs implementation - A new `cogs` folder in the `.agentforge` folder for Cog YAML files - Started Documentation for Cogs - def not final.
- Added the ability to references nested placeholders in prompt templates, this was crucial to keeping the Cog implementation clean and simple. - Reworked most of the implementation for Cogs, we now have a very basic POC. - Added validation methods to separate loading and validation from the actual Cog execution, simplifying implementation. - Reworked the tests for Cogs, they are currently passing, more unit test and real test case pending.
Renamed a few methods and variables for consistence and abstracted a new internal execute flow method to make the code easier to read and follow.
Finished the basic Cog Implementation and made a basic cog guide, more documentation changes pending.
Cogs can now automatically parse agent responses based on a defined format. Currently supported formats are: - xml - json - yaml - ini - csv - markdown
Some work on memory done.
The base memory class is ready for integration with the Cog class.
A very basic and crude integration of memory into cogs, it needs work and polish but every major component is in place
Apparently gitkraken didn't catch the changes to index fast enough?
Moves logger initialization to the TrailRecorder class itself, simplifying its instantiation and removing the need to pass a logger instance during initialization. This change promotes encapsulation and reduces dependencies.
Enhances code clarity and maintainability by refining log levels and message formats. - Streamlines memory management with explicit query/update triggers, improving agent-memory interactions. - Corrects logging calls to use appropriate levels (debug, info, warning, error, critical) for better clarity and filtering. - Implements more informative debugging and error messaging, aiding in issue identification and resolution.
Forcing naming convention on all files and folders
- Introduced a @Property 'settings' on Config for elegant, dot-notation access to all framework settings as a dataclass. - Refactored logger, storage, actions, and persona memory to use config.settings instead of raw dicts. - Updated prompt processor to handle both dict and dataclass for persona settings, fixing test suite breakage. - All tests pass. Because who needs .get('foo', 'bar') when you can have dot notation and a little peace of mind?
Cleanup for readability and maintainability
Quick update to Update Assistant
Refactored MemoryManager for clarity and maintainability
Refactored TransitionResolver for clarify and maintainability
Adds chat history functionality to Cogs, allowing them to retain and utilize conversation history. Introduces ChatHistoryMemory class for managing chat interactions. Configures Cog to load and record chat history, enabling context-aware agent responses. Refactors prompt configurations for improved agent interaction with chat history.
Enhances chat history memory by sorting records chronologically and formatting the output for better readability. Sorts history entries by timestamp, falling back to ID if timestamp is unavailable. Streamlines data presentation in the scratchpad.
Removes the episodic memory, journal, and memory tools modules. These modules are not ready for release and will be reintegrated in future versions.
Updates and consolidates example cogs configurations for clarity and better memory management. Refactors prompts for improved user interaction and agent performance.
Updated Test Suite to new functionality.
Adds JSON preprocessing to handle LLM output variations, improving parsing success. Allows for dot-notated 'end' values in cog transitions, enabling the retrieval of nested fields from agent outputs. Updates config parsing to properly evaluate 'end' conditions in transitions.
Documentation updated to match AF v0.6.0
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
v0.6.0 Changelog
Major Release: Focus on Cognitive Architectures (Cogs)
Highlights
Cogs: Cognitive Architectures Implemented
Memory System Overhaul
MemoryManagerfor improved clarity and maintainability.Agent and Persona Enhancements
TrailRecorderandThoughtTrailEntry.Infrastructure and Documentation
Notable Breaking Changes
This release marks a significant step forward in agent modularity, memory integration, and cognitive architecture flexibility, laying the groundwork for more advanced, context-aware AI systems.