openai/codex
1
0
Fork 0
mirror of https://github.com/openai/codex.git synced 2026-11-06 20:20:34 +00:00
Code Issues Packages Projects Releases Wiki Activity

Show realtime text in TUI transcript

Browse Source 
Request text-only output from realtime v2, render realtime transcript deltas in the TUI transcript, and clean common text response wrappers before display.

Co-authored-by: Codex <noreply@openai.com>
This commit is contained in:
Ahmed Ibrahim
2026-04-13 12:57:57 -07:00
co-authored by Codex
parent d905376628
commit 8a122f64fd
8 changed files with 239 additions and 68 deletions
Download Patch File Download Diff File Expand all files Collapse all files
codex-rs/codex-api/src/endpoint/realtime_websocket/methods.rs
+2 -12
View File
@@ -1519,7 +1519,7 @@ mod tests {
first_json["session"]["type"],
Value::String("realtime".to_string())
);
assert_eq!(first_json["session"]["output_modalities"], json!(["audio"]));
assert_eq!(first_json["session"]["output_modalities"], json!(["text"]));
assert_eq!(
first_json["session"]["audio"]["input"]["format"],
json!({
@@ -1541,17 +1541,7 @@ mod tests {
"create_response": true,
})
);
assert_eq!(
first_json["session"]["audio"]["output"]["format"],
json!({
"type": "audio/pcm",
"rate": 24_000,
})
);
assert_eq!(
first_json["session"]["audio"]["output"]["voice"],
Value::String("cedar".to_string())
);
assert!(first_json["session"]["audio"].get("output").is_none());
assert_eq!(
first_json["session"]["tools"][0]["type"],
Value::String("function".to_string())
codex-rs/codex-api/src/endpoint/realtime_websocket/methods_v2.rs
+5 -13
View File
@@ -14,8 +14,6 @@ use crate::endpoint::realtime_websocket::protocol::RealtimeVoice;
use crate::endpoint::realtime_websocket::protocol::SessionAudio;
use crate::endpoint::realtime_websocket::protocol::SessionAudioFormat;
use crate::endpoint::realtime_websocket::protocol::SessionAudioInput;
use crate::endpoint::realtime_websocket::protocol::SessionAudioOutput;
use crate::endpoint::realtime_websocket::protocol::SessionAudioOutputFormat;
use crate::endpoint::realtime_websocket::protocol::SessionFunctionTool;
use crate::endpoint::realtime_websocket::protocol::SessionNoiseReduction;
use crate::endpoint::realtime_websocket::protocol::SessionToolType;
@@ -25,10 +23,10 @@ use crate::endpoint::realtime_websocket::protocol::SessionUpdateSession;
use crate::endpoint::realtime_websocket::protocol::TurnDetectionType;
use serde_json::json;
const REALTIME_V2_OUTPUT_MODALITY_AUDIO: &str = "audio";
const REALTIME_V2_OUTPUT_MODALITY_TEXT: &str = "text";
const REALTIME_V2_TOOL_CHOICE: &str = "auto";
const REALTIME_V2_BACKGROUND_AGENT_TOOL_NAME: &str = "background_agent";
const REALTIME_V2_BACKGROUND_AGENT_TOOL_DESCRIPTION: &str = "Send a user request to the background agent. Use this as the default action. If the background agent is idle, this starts a new task and returns the final result to the user. If the background agent is already working on a task, this sends the request as guidance to steer that previous task. If the user asks to do something next, later, after this, or once current work finishes, call this tool so the work is actually queued instead of merely promising to do it later.";
const REALTIME_V2_BACKGROUND_AGENT_TOOL_DESCRIPTION: &str = "Send a user request to the background agent. Use this as the default action. Do not rephrase the user's ask or rewrite it in your own words; pass along the user's own words. If the background agent is idle, this starts a new task and returns the final result to the user. If the background agent is already working on a task, this sends the request as guidance to steer that previous task. If the user asks to do something next, later, after this, or once current work finishes, call this tool so the work is actually queued instead of merely promising to do it later.";
pub(super) fn conversation_item_create_message(text: String) -> RealtimeOutboundMessage {
RealtimeOutboundMessage::ConversationItemCreate {
@@ -59,7 +57,7 @@ pub(super) fn conversation_handoff_append_message(
pub(super) fn session_update_session(
instructions: String,
session_mode: RealtimeSessionMode,
voice: RealtimeVoice,
_voice: RealtimeVoice,
) -> SessionUpdateSession {
match session_mode {
RealtimeSessionMode::Conversational => SessionUpdateSession {
@@ -67,7 +65,7 @@ pub(super) fn session_update_session(
r#type: SessionType::Realtime,
model: None,
instructions: Some(instructions),
output_modalities: Some(vec![REALTIME_V2_OUTPUT_MODALITY_AUDIO.to_string()]),
output_modalities: Some(vec![REALTIME_V2_OUTPUT_MODALITY_TEXT.to_string()]),
audio: SessionAudio {
input: SessionAudioInput {
format: SessionAudioFormat {
@@ -83,13 +81,7 @@ pub(super) fn session_update_session(
create_response: true,
}),
},
output: Some(SessionAudioOutput {
format: Some(SessionAudioOutputFormat {
r#type: AudioFormatType::AudioPcm,
rate: REALTIME_AUDIO_SAMPLE_RATE,
}),
voice,
}),
output: None,
},
tools: Some(vec![SessionFunctionTool {
r#type: SessionToolType::Function,