-
Notifications
You must be signed in to change notification settings - Fork 1.3k
Enhance on_tool_start
Hook to Include Tool Call Arguments
#252
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
Until the linked PR is accepted, is there any other way to obtain the tool call arguments using hooks before the agent calls the tool? |
Some workaround I think of is either using LLM tracing, put logging in tool implement or hijack the network call |
@MrJarnould @vanhtuan0409 Yes, I do it in the tool calling internal for now. But it's more reasonable to put this in the tracing hook :) |
I hope your PR gets reviewed soon! |
Any idea when this enhancement will be released ? |
@rm-openai Could you please take a look at this issue? Thanks! |
Please read this first
Describe the feature
Currently, the
on_tool_start
hook in the Agents SDK only provides access to the Agent, Tool (definition), and context, but does not include the actual tool call arguments.This limitation makes it difficult to trace and debug the agent's lifecycle, as there's no way to inspect the exact parameters passed to the tool before execution.
Proposed Enhancement:
Modify the
on_tool_start
hook definition to include the tool call arguments, providing full visibility into the tool invocation process.Why This Improvement Matters:
The text was updated successfully, but these errors were encountered: