Skip to content

fix: render multimodal content parts in span attributes#3

Open
Munksgaard wants to merge 1 commit intolostbean:mainfrom
Munksgaard:fix/render-multimodal-content-parts
Open

fix: render multimodal content parts in span attributes#3
Munksgaard wants to merge 1 commit intolostbean:mainfrom
Munksgaard:fix/render-multimodal-content-parts

Conversation

@Munksgaard
Copy link
Copy Markdown

extract_text_content/1 only handled :text parts; file/image/audio parts collapsed to empty strings, making multimodal LLM inputs invisible in traces.

Now renders them as '[file: application/pdf, 1024 bytes]' placeholders by default. Set config :agent_obs, include_multimodal_data: true to inline the raw base64 data instead — useful for replay/playground in dev.

extract_text_content/1 only handled :text parts; file/image/audio parts
collapsed to empty strings, making multimodal LLM inputs invisible in traces.

Now renders them as '[file: application/pdf, 1024 bytes]' placeholders by
default. Set 'config :agent_obs, include_multimodal_data: true' to inline
the raw base64 data instead — useful for replay/playground in dev.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant