![LaVague Update #7](/content/images/size/w2640/2024/07/thumbnail.png)
LaVague Update #7
Welcome to our weekly updates, where we announce the latest news on our open-source framework for Web Agents & share what's coming next and how you can get involved!
Welcome to our 7th dev update.
Over the past couple of weeks, we have:
- Improved overall performance & can now handle
iframes
and multiple tabs - 🔧 Added some basic debugging tools
- 🧪 Added a LaVague test runner tool
- 🚀 Switched action objects to
yaml
format to gain a small performance boost + added extractor functions to extract generatedyaml
,json
&Python
code - 🪙 Documented the TokenCounter module
🏅 Community PRs merged:
- Option to log to a SQL database added by GitHub user jashuajoy
TokenCounter
tool added byjashuajoy
- Fix for issues with docs formatting by rmrf
📢 Help wanted:
- Help us to create a custom file upload/download action
- Rewriting of the agent to have better variable and output management
This is just a selection of help-wanted
issues, see the full list here.
🛤️ Coming soon:
- More performance improvements!
- Launch of the
LaVague chrome extension
- Improvements to the dataset pipeline with local HTML for reproducible actions
- Documentation updates & additions
- Handling of ambiguous elements by adding expanded context where needed
- Add LLM output format validation for
json
,Python
&yaml
outputs - Start of work on an
agent hub
for sharing agents
If you have any questions, feedback or want to see how you can get involved in the project, join us on Discord.