fix(hook): trace subagents launched by subagents

[pushkar]

Subagents launched by another subagent were never traced. This passes the subagent transcript map down so they nest under the agent that launched them.

Problem

emit_subagent_observations walked an agent's turns with subagent_transcripts_by_tool_use_id=None. An Agent call inside a subagent therefore got a Tool: Agent span but no span of its own: none of its generations or tool calls were emitted, even though its transcript (spawnDepth: 2) is on disk and already in the map.

Fix

• emit_subagent_observations takes the map and hands it to emit_turn_observations for the agent's turns.
• Both regular-subagent call sites pass it in: synchronous agents and deferred async ones.
• Workflow agents are unchanged.

Testing

• test_nested_subagents_document_current_non_recursive_emission_behavior asserted the gap. It's now test_nested_subagents_emit_inner_agent_under_outer_agent and requires the inner agent's span under the outer agent, with a generation under it. It fails before the fix and passes after.
• uv run --group dev pytest: 279 passed, also on Python 3.10. ruff check --select F,E9 is clean.
• Live run against a self-hosted Langfuse: a parent agent that starts a nested agent now traces as Subagent: <inner> ← Subagent: <outer> ← Conversational Turn, including the inner agent's Bash call.

[claude]

Claude Code Review

This pull request is from a fork — automated review is disabled. A repository maintainer can comment @claude review to run a one-time review.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}