feat: support dynamic max_steps via environment variable

fix: complete code replacement to eliminate import error

fix: modify prompt templates and add web tools

chore: change default model to openrouter free llama3.1

fix: add fake gpu function to bypass startup check without pickle error

fix: remove spaces.gpu and initialize agent inside call to avoid pickle error

fix: add ddgs dependency for search tool

fix: install litellm extra for smolagents

fix: add @spaces.GPU decorator for Zero GPU environment

fix: remove hardcoded token and use env var

update my agent with token

Update app.py

Update app.py

Update app.py

Update app.py

Update requirements.txt

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update README.md

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update README.md

Update app.py

Update app.py

Update app.py

Update app.py

Update requirements.txt

Update app.py

Update requirements.txt

Update requirements.txt

Create app.py

Create requirements.txt

initial commit