Python: Input & output attributes in invoke_agent span

[TaoChenOSU]

Motivation and Context

The OTel workgroup has been thinking about enhancing tracing for agents. This is a proposal from our tracing team: open-telemetry/semantic-conventions#2528. This PR implements part of the proposal so that when the proposal gets merged, we can quickly follow.

Description

Two new attributes are added to the invoke_agent span, namely the gen_ai.agent.invocation_input and gen_ai.agent.invocation_output.

Contribution Checklist

• The code builds clean without any errors or warnings
• The PR follows the SK Contribution Guidelines and the pre-submission formatting script raises no violations
• All unit tests pass, and I have added new tests where possible
• I didn't break anyone 😄

[moonbox3]

Python Test Coverage Report •

File	Stmts	Miss	Cover	Missing
agents/azure_ai
azure_ai_agent.py	362	180	50%	99, 101–104, 106–108, 110–115, 117, 119, 121–123, 125, 142, 144–147, 149–154, 156, 161–162, 167–170, 175–179, 181–184, 186–192, 197, 199–202, 204–206, 208–211, 213, 215, 225–226, 228–231, 233–234, 239, 241–247, 249, 282, 293–294, 299–300, 303, 308–313, 320–321, 323, 328–329, 340–346, 398, 400, 406, 417, 420, 422, 424, 426, 460, 491–493, 495, 497–498, 500–503, 506–508, 510–511, 514, 517–526, 528, 537–538, 541–543, 545–546, 556–557, 559, 577, 579, 582, 584–585, 587–588, 590, 599–600, 602, 604–607, 610, 612, 615–617, 621, 690, 796, 900, 973
agents/bedrock
bedrock_agent.py	254	56	77%	78, 84–85, 94, 160, 287, 300, 319, 335–338, 344, 346, 349, 354, 386, 389, 402, 434–435, 442–445, 453–454, 458, 491, 494, 507, 526–529, 531–534, 576, 587–588, 600, 636–637, 647, 658, 675–676, 688, 705, 707, 709, 711–712, 714
agents/chat_completion
chat_completion_agent.py	230	33	85%	69, 74, 79, 85, 101, 103, 109–113, 171, 174, 184, 192, 195, 197, 206, 232, 251, 253–254, 256–257, 260–262, 264, 266, 268, 416, 513, 590
agents/copilot_studio
copilot_studio_agent.py	247	247	0%	3–8, 10–14, 16–19, 23–26, 31–36, 42, 50, 56–60, 63–64, 67, 78–84, 86–87, 89, 93–95, 99–102, 104, 106, 108, 110–111, 115–119, 121–122, 129–131, 135–137, 139–140, 143, 148, 151, 153, 156–157, 162–163, 168, 170–175, 177–180, 182–184, 193–194, 197, 208–210, 212–213, 215–216, 218, 220–221, 223, 225–227, 229, 231–232, 234–236, 238–245, 247–248, 255–256, 259–260, 262, 287–288, 290, 295–302, 304–305, 310–311, 315–316, 318, 320, 322–324, 326–327, 364–365, 367–368, 383–384, 386–388, 390, 394, 405, 409–411, 432, 438–439, 441, 443–444, 452, 454, 456–458, 481, 487–488, 490, 492, 500, 502–504, 535, 541–542, 544, 546–547, 555, 557–559, 565, 574–575, 577, 579–580, 582, 584–586, 588–589, 591, 593, 595–597, 602–604, 607, 609, 611–612, 614–618, 621, 623–624, 626–627, 629–630, 632–635, 637–638, 644, 652–653, 658
agents/open_ai
openai_assistant_agent.py	363	71	80%	75, 111–114, 119, 123–124, 157, 174–175, 183–188, 195–196, 199, 204–205, 216–222, 400–401, 408–409, 411–412, 414–415, 417–422, 424, 472–473, 476, 479, 482–491, 493, 503, 527–528, 566, 582, 592, 608, 621, 657, 801, 910, 1017
openai_responses_agent.py	410	104	74%	66, 95, 114–117, 125, 129–130, 162–166, 170, 175, 180, 185, 191, 196–197, 200, 203, 208–213, 218–219, 221–222, 228–231, 237, 239–240, 244–248, 339, 343, 347, 349, 355, 368, 370, 372, 376, 471–472, 479–480, 482–483, 485–486, 488–493, 495, 543–544, 547, 592, 608, 618, 652, 660, 663, 667, 669, 672, 695, 698–702, 712, 752, 777, 788, 879, 999, 1040, 1042, 1116, 1156–1157, 1159–1163, 1182
utils/telemetry/agent_diagnostics
decorators.py	116	8	93%	163, 230–235, 240
gen_ai_attributes.py	7	0	100%
TOTAL	26775	4601	82%

Python Unit Test Overview

Tests	Skipped	Failures	Errors	Time
3661	22 💤	0 ❌	0 🔥	1m 56s ⏱️