apunkt/plano
1
0
Fork 0
mirror of https://github.com/katanemo/plano.git synced 2026-05-06 22:32:42 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 3

Add support for Amazon Bedrock Converse and ConverseStream (#588)

Browse source 
* first commit to get Bedrock Converse API working. Next commit support for streaming and binary frames

* adding translation from BedrockBinaryFrameDecoder to AnthropicMessagesEvent

* Claude Code works with Amazon Bedrock

* added tests for openai streaming from bedrock

* PR comments fixed

* adding support for bedrock in docs as supported provider

* cargo fmt

* revertted to chatgpt models for claude code routing

---------

Co-authored-by: Salman Paracha <salmanparacha@MacBook-Pro-288.local>
Co-authored-by: Adil Hafeez <adil.hafeez@gmail.com>
This commit is contained in:
Salman Paracha 2025-10-22 11:31:21 -07:00 • committed by GitHub
parent ba826b1961
commit 9407ae6af7
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
35 changed files with 7362 additions and 1493 deletions
Download patch file Download diff file Expand all files Collapse all files

19
crates/brightstaff/src/handlers/chat_completions.rs
View file

@ -1,7 +1,8 @@
use bytes::Bytes;
use common::configuration::{ModelAlias, ModelUsagePreference};
use common::consts::ARCH_PROVIDER_HINT_HEADER;
use common::consts::{ARCH_IS_STREAMING_HEADER, ARCH_PROVIDER_HINT_HEADER};
use hermesllm::apis::openai::ChatCompletionsRequest;
use hermesllm::clients::endpoints::SupportedUpstreamAPIs;
use hermesllm::clients::SupportedAPIs;
use hermesllm::{ProviderRequest, ProviderRequestType};
use http_body_util::combinators::BoxBody;
@ -56,6 +57,7 @@ pub async fn chat(
// Model alias resolution: update model field in client_request immediately
// This ensures all downstream objects use the resolved model
let model_from_request = client_request.model().to_string();
let is_streaming_request = client_request.is_streaming();
let resolved_model = if let Some(model_aliases) = model_aliases.as_ref() {
if let Some(model_alias) = model_aliases.get(&model_from_request) {
debug!(
@ -84,10 +86,16 @@ pub async fn chat(
let chat_completions_request_for_arch_router: ChatCompletionsRequest =
match ProviderRequestType::try_from((
client_request,
&SupportedAPIs::OpenAIChatCompletions(hermesllm::apis::OpenAIApi::ChatCompletions),
&SupportedUpstreamAPIs::OpenAIChatCompletions(
hermesllm::apis::OpenAIApi::ChatCompletions,
),
)) {
Ok(ProviderRequestType::ChatCompletionsRequest(req)) => req,
Ok(ProviderRequestType::MessagesRequest(_)) => {
Ok(
ProviderRequestType::MessagesRequest(_)
| ProviderRequestType::BedrockConverse(_)
| ProviderRequestType::BedrockConverseStream(_),
) => {
// This should not happen after conversion to OpenAI format
warn!("Unexpected: got MessagesRequest after converting to OpenAI format");
let err_msg = "Request conversion failed".to_string();
@ -190,6 +198,11 @@ pub async fn chat(
header::HeaderValue::from_str(&model_name).unwrap(),
);
request_headers.insert(
header::HeaderName::from_static(ARCH_IS_STREAMING_HEADER),
header::HeaderValue::from_str(&is_streaming_request.to_string()).unwrap(),
);
if let Some(trace_parent) = trace_parent {
request_headers.insert(
header::HeaderName::from_static("traceparent"),