Expose input messages to BeforeInvocationEvent #1007
Conversation
athewsey
had a problem deploying
to
manual-approval
GitHub Actions
Failure
athewsey
changed the title
Add incoming messages to BeforeInvocationEvent so this hook can be used for guardrails aiming to check (and perhaps redact) user inputs before kicking off the event loop.
athewsey
force-pushed
the
feat/beforeinvocation
branch
from
23f6080 to
3d72200
Compare
athewsey
had a problem deploying
to
manual-approval
GitHub Actions
Failure
mkmeral
left a comment
mkmeral
left a comment
There was a problem hiding this comment.
Thanks for the PR, and I'm sorry for the delayed response.
The feature and PR makes a lot of sense, I just had 2 small comments.
| self.hooks.invoke_callbacks(BeforeInvocationEvent(agent=self)) | ||
| if not self.messages and not prompt: | ||
| raise ValueError("No conversation history or prompt provided") | ||
| temp_messages: Messages = self.messages + self._convert_prompt_to_messages(prompt) |
There was a problem hiding this comment.
can we just send the new messages? I think that would allow devs to differentiate between new messages (what we pass), and agent.messages (what already exists in history).
| with self.tracer.tracer.start_as_current_span( | ||
| "execute_structured_output", kind=trace_api.SpanKind.CLIENT | ||
| "execute_structured_output", | ||
| attributes={ |
There was a problem hiding this comment.
do we need to update the span?
|
Apologies for the delay on this issue, but we have made some significant changes since this was submitted. We need to rebase this pr in order to fit with the current state of the codebase. |
|
Hey @athewsey! 👋 I noticed this PR has merge conflicts. The Since this is a valuable feature (tagged as The feature makes a lot of sense for implementing early guardrails on user input before it gets added to memory. I saw @mkmeral's feedback from Jan 8th - looks like the PR just needs:
Would love to see this land! 🚀 Comment from strands-coder autonomous agent 🦆 |
4 participants
Description
Add the incoming
messagestoBeforeInvocationEvent, so that this hook can be used for implementing guardrails that check (and perhaps redact) user inputs before kicking off the event loop or raisingMessageAddedEvents.Related Issues
#1006
Documentation PR
None yet - Should be some small updates required if we go ahead. I can work on this if reviewers are OK with the change in principle?
Type of Change
New feature
Testing
Updated existing unit tests to validate the messages are passed through correctly (both in structured output and normal agent invocation).
hatch run prepareChecklist
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.