feat(observability): trace skill execution

[zakahan]

Summary

• add a shared Skill observability model for legacy VeADK SkillsTool and Google ADK SkillToolset operations
• propagate W3C trace context through InvokeTool into the sandbox python agent process
• attribute real LLM input/output token usage to the active Skill without high-cardinality metric labels
• add Skill invocation, error, duration, and token metrics while retaining the legacy latency metric
• prevent active Skill state from leaking across invocation IDs

Validation

• 30 focused tests passed
• Ruff 0.11.12 lint and format checks passed
• branch is based on this fork latest origin/main

Notes

Only traceparent and tracestate cross the sandbox boundary; baggage and prompt/output content are not added by this change.

[github-actions]

license-eye has totally checked 620 files.

Valid	Invalid	Ignored	Fixed
619	1	0	0

Click to see the invalid file list

• tests/tracing/test_skill_observability.py

[github-actions]

Suggested change

	from types import SimpleNamespace
	# Copyright (c) 2025 Beijing Volcano Engine Technology Co., Ltd. and/or its affiliates.
	#
	# Licensed under the Apache License, Version 2.0 (the "License");
	# you may not use this file except in compliance with the License.
	# You may obtain a copy of the License at
	#
	# http://www.apache.org/licenses/LICENSE-2.0
	#
	# Unless required by applicable law or agreed to in writing, software
	# distributed under the License is distributed on an "AS IS" BASIS,
	# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
	# See the License for the specific language governing permissions and
	# limitations under the License.
	from types import SimpleNamespace