OTEL logging integration & System-wise Traces

[toubatbrian]

Add tracing and structured logging to agents-js with 100% Python parity.

Breakdown

✅ Implemented in TypeScript: 20/20 Active Spans

Span Name	Python	TypeScript	Status
job_entrypoint	job.py:739	job_proc_lazy_main.ts:105	✅
llm_request	llm.py:177	llm.ts:192	✅
llm_request_run	llm.py:~184	llm.ts:147	✅
llm_node	generation.py:86	generation.ts:475	✅
tts_request (streaming)	tts.py:367	tts.ts:217	✅
tts_request (non-streaming)	tts.py:241	tts.ts:478	✅
tts_request_run	tts.py:~380,~254	tts.ts:171,432	✅
tts_node	generation.py:214	generation.ts:531	✅
agent_session	agent_session.py:519	agent_session.ts:322	✅
agent_turn	agent_activity.py:1673	agent_activity.ts:1654	✅
agent_turn (realtime)	agent_activity.py:2086	(realtime flow)	✅
user_turn	audio_recognition.py:660	audio_recognition.ts:571	✅
function_tool	generation.py:544	generation.ts:859	✅
agent_speaking	agent_session.py:1159	agent_session.ts:515	✅
user_speaking	agent_session.py:1190	agent_session.ts:554	✅
start_agent_activity	agent_activity.py:425	agent_activity.ts:210	✅
drain_agent_activity	agent_activity.py:577	agent_activity.ts:2167	✅
on_enter	agent_activity.py:446	agent_activity.ts:304	✅
on_exit	agent_activity.py:584	agent_activity.ts:2183	✅
eou_detection	audio_recognition.py:~493	audio_recognition.ts:366	✅

---

⏸️ Python-Specific Features (Not in TS): 2

Span Name	Python	Reason
resume_agent_activity	agent_activity.py:635	Pause/resume not in TS
pause_agent_activity	agent_activity.py:656	Pause/resume not in TS

Pause/resume functionality doesn't exist in TypeScript agent_activity yet.

N/A (Feature Not in TypeScript): 1

Span Name	Python	Reason
judge_evaluation	run_result.py:829	Judge feature not implemented in TS

📈 Span Summary:
  - start_agent_activity: 1
  - on_enter: 1
  - job_entrypoint: 1
  - user_speaking: 4
  - eou_detection: 3
  - user_turn: 3
  - llm_request_run: 5
  - llm_request: 5
  - llm_node: 5
  - tts_node: 5
  - agent_speaking: 3
  - agent_turn: 5
  - function_tool: 2
  - on_exit: 1
  - drain_agent_activity: 1
  - agent_session: 1

🌳 Span Hierarchies:

Trace 1 (job_entrypoint):
└─ job_entrypoint (5474.11ms)
   └─ agent_session (55046.00ms)
      ├─ start_agent_activity (3.33ms)
         └─ on_enter (2.76ms)
      ├─ user_speaking (1469.25ms)
      ├─ user_turn (2091.39ms)
      │  transcript: "Hello can you hear me."
      ├─ agent_turn (6292.53ms)
      │  speech_id: speech_a6409301-17d
      │  user_input: "Hello can you hear me."
         ├─ llm_node (2142.84ms)
            └─ llm_request (2138.71ms)
               model: openai/gpt-4.1-mini
               tokens: 94→17
               └─ llm_request_run (2138.31ms)
         └─ tts_node (3628.79ms)
      ├─ eou_detection (106.69ms)
      ├─ agent_speaking (3715.15ms)
      ├─ user_speaking (2561.00ms)
      ├─ user_turn (4942.13ms)
      │  transcript: "What is the weather only san francisco."
      ├─ user_speaking (1444.43ms)
      ├─ agent_turn (2596.05ms)
      │  speech_id: speech_4960de9e-9a5
      │  user_input: "What is the weather only san francisco."
         ├─ llm_node (2195.47ms)
            └─ llm_request (2194.23ms)
               model: openai/gpt-4.1-mini
               tokens: 127→15
               └─ llm_request_run (2194.12ms)
         ├─ tts_node (2195.61ms)
         └─ function_tool (501.37ms)
            tool: getWeather
      ├─ eou_detection (36.99ms)
      ├─ agent_turn (6425.45ms)
      │  speech_id: speech_7b4ef0c0-26e
         ├─ llm_node (1407.95ms)
            └─ llm_request (1406.97ms)
               model: openai/gpt-4.1-mini
               tokens: 162→16
               └─ llm_request_run (1406.73ms)
         └─ tts_node (3148.83ms)
      ├─ agent_speaking (4502.66ms)
      ├─ user_speaking (1627.59ms)
      ├─ user_turn (2578.55ms)
      │  transcript: "What's the time."
      ├─ agent_turn (2073.87ms)
      │  speech_id: speech_8cb6d87e-726
      │  user_input: "What's the time."
         ├─ llm_node (2072.98ms)
            └─ llm_request (2071.19ms)
               model: openai/gpt-4.1-mini
               tokens: 189→11
               └─ llm_request_run (2071.10ms)
         ├─ tts_node (2072.82ms)
         └─ function_tool (0.26ms)
            tool: getCurrentTime
      ├─ eou_detection (28.43ms)
      ├─ agent_turn (5553.00ms)
      │  speech_id: speech_76a71364-ef7
         ├─ llm_node (2407.05ms)
            └─ llm_request (2405.40ms)
               model: openai/gpt-4.1-mini
               tokens: 217→11
               └─ llm_request_run (2405.34ms)
         └─ tts_node (3532.38ms)
      ├─ agent_speaking (2648.94ms)
      └─ drain_agent_activity (0.30ms)
         └─ on_exit (0.15ms)

[changeset-bot]

⚠️ No Changeset found

Latest commit: 780d42e

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR