Dirk Brand
Dirk Brand
## Description **Please include:** - This adds the replicate toolkit for generating videos and images ## Type of change Please check the options that are relevant: - [ ] Bug...
## Description - Add multi-modal support for agents. Image and Video as model output can be added to agents. - Add toolkits for Replicate, FAL and luma ## Type of...
## Description Fixes #1556 Crawl4ai has deprecated their synchronous webcrawler. This switches to using the AsyncWebCrawler. In future when we support async tool calling, we can use it more directly....
## Description This adds the GiphyTools and examples of how to search gifs as an agent. ## Type of change Please check the options that are relevant: - [ ]...
## Description This implements `response_audio` to get real-time audio responses from models directly. This bundles all of the audio responses together into the final run response. ## Type of change...
## Description This adds a workflow to showcase a simple HTML 5 game generator. Followups will include sounds and sprites for the game generator. ## Type of change Please check...
## Description A complete revamp of teams. TBD
## Description Users have to import unnecessary dependencies if they only want one or the other. This change unfortunately cannot be made non-breaking, because the import has to be removed....
## Description This addresses a confusing DX point where users often have to set response_model and structured_output. This changes the behaviour to default to structured output if supported. json_response_mode can...
## Description This will make all vector DBs async (where possible) as well as all readers --- ## Type of change Please check the options that are relevant: - [...