Agent Project - SimpleDSPy Agent

This project features a DSPy-based agent that can run commands and interact with users. It uses OpenRouter models for processing requests.

Setup

Ensure you are in the project root directory: /home/tom/git/agent/
Create and activate a Python virtual environment (if you haven't already for this project):
```
python3.11 -m venv .venv
source .venv/bin/activate
```
Install or update dependencies: If requirements.txt is new or you've just created it:
```
uv pip install -r requirements.txt
```
If you are adding to an existing requirements.txt, ensure the new packages are installed.
Set up OpenRouter API key:
- Create a free account at OpenRouter
- Add your API key to environment variables:
```
echo "export OPENROUTER_API_KEY='your_api_key'" >> ~/.bashrc
source ~/.bashrc
```
Run the SimpleDSPy agent:
```
python agent_simpledspy.py
```

The agent's performance depends on quality training data stored in .simpledspy/modules/. To improve results:

Review logged interactions:
```
less .simpledspy/modules/*/logged.jsonl
```
Add good samples to training data: Copy well-performing examples from logged.jsonl to corresponding training.jsonl files
Correct problematic samples: Edit training.jsonl to fix incorrect responses
Add new training examples: Create new entries in training.jsonl for desired behaviors

The main agent script is agent_simpledspy.py
Training data is stored in .simpledspy/modules/
Use --lm flag to switch models: flash, r1, or dv3
```
python agent_simpledspy.py --lm r1
```

Name		Name	Last commit message	Last commit date
Latest commit History 122 Commits
agent_repl		agent_repl
dspy_programs		dspy_programs
tests		tests
.gitignore		.gitignore
README.md		README.md
active_learning_loop.py		active_learning_loop.py
agent_simpledspy.py		agent_simpledspy.py
coding_agent_repl.py		coding_agent_repl.py
deepseek_simba_optimizer.py		deepseek_simba_optimizer.py
generator_module.py		generator_module.py
info.md		info.md
memory_gan_optimizer.py		memory_gan_optimizer.py
notes.md		notes.md
online_optimization_system.py		online_optimization_system.py
requirements.txt		requirements.txt
shell_wrapper.py		shell_wrapper.py
simpledspy_demo.py		simpledspy_demo.py
simpledspy_main.py		simpledspy_main.py
taskwarrior_dspy_agent.py		taskwarrior_dspy_agent.py
taskwarrior_dspy_definitions.py		taskwarrior_dspy_definitions.py
test_shell_wrapper.py		test_shell_wrapper.py
value_network.py		value_network.py