mirror of
https://github.com/humanlayer/humanlayer.git
synced 2025-08-20 19:01:22 +03:00
a3d286a7a7
What I did ----------- - Enabled state preservation in `FunctionCallSpec` and `HumanContactSpec`, allowing applications to maintain context across request lifecycles without storing state. - Updated documentation and examples to demonstrate the new state preservation feature. How I did it ----------- - Added an optional `state` field to the models. - Created new examples and updated existing ones to showcase state usage. - Wrote tests to verify state is preserved across requests. - Refactored session management in `async_cloud.py` for simplicity. - [x] I have ensured `make check test` passes How to verify it ----------- - Run `make check test` to ensure all tests pass. - Review the `app-statehooks.py` example for state preservation without webhooks. - Check the updated documentation under the **State Preservation** section. Description for the changelog -------------- - Add state preservation support to `FunctionCallSpec` and `HumanContactSpec`, enabling applications to maintain context across requests without storing state.
Human Layer Curl Example
This example demonstrates how to use Human Layer's API with simple curl commands to create and check function call approvals.
Prerequisites
curljq(for JSON parsing)- Valid
HUMANLAYER_API_KEYenvironment variable set with your API key from Human Layer
Basics
to create an approval request, run a curl like:
curl -X POST https://api.humanlayer.dev/humanlayer/v1/function_calls \
-H "Authorization: Bearer ${HUMANLAYER_API_KEY}" \
-H "Content-Type: application/json" \
-d '{
"run_id": "curl-example",
"call_id": "1234-5678-...",
"spec": {
"fn": "send_email",
"kwargs": {
"to": "user@example.com",
"subject": "Hello",
"body": "This is a test email"
}
}
}'
Usage w/ Makefile
Test Large Context Size Limit
Test the context size limit by sending a large payload:
export HUMANLAYER_API_KEY=<your-api-key>
make test-large-context
This will send a function call with a large array in the state field to test the context size limits.
1. Create an Approval Request
Create a new function call approval request:
export HUMANLAYER_API_KEY=<your-api-key>
make create-approval
This will create a new approval request with a randomly generated UUID. The output will look like:
{
"run_id": "curl-example",
"call_id": "1234-5678-...",
"spec": {
"fn": "send_email",
"kwargs": {
"to": "user@example.com",
"subject": "Hello",
"body": "This is a test email"
}
},
"status": {
"approved": false,
"comment": null
}
}
2. Check Approval Status
Check the status of your approval request, using the call_id from the approval request:
make check-approval random_id=1234-5678-...
If not yet approved, you'll see:
{
"run_id": "curl-example",
"call_id": "1234-5678-...",
"spec": {
"fn": "send_email",
"kwargs": {
"to": "user@example.com",
"subject": "Hello",
"body": "This is a test email"
}
},
"status": {
"approved": false,
"comment": null
}
}
Once approved:
{
"run_id": "curl-example",
"call_id": "1234-5678-...",
"spec": {
"fn": "send_email",
"kwargs": {
"to": "user@example.com",
"subject": "Hello",
"body": "This is a test email"
}
},
"status": {
"approved": true,
"comment": "Approved by reviewer"
}
}
3. Run Complete Flow
To create a request and wait for approval:
make run-agent
This will:
- Create an approval request
- Poll for approval status every 3 seconds
- Output the approval comment once approved
Example output while waiting:
waiting for approval
waiting for approval
waiting for approval
Once approved:
approval granted with comment "Approved by reviewer"