• Adds instrumentation for tracking LLM input and output token usage
• Emits a telemetry event name ai_completion when an LLM tool call loop finishes

Note that the this ai_completion event is different from the existing ai_message event:

• ai_message: one event per user/LLM/tool/progress message; use this to understand conversation length/actions
• ai_completion: one event per LLM interaction; use this to understand token usage

Checklist:

• Covered by tests
• Ran it and it works as intended
• Reviewed the diff before requesting a review
• Checked for unhandled edge cases
• Linked the issues it closes
• Checked if the docs need to be updated. If so, create a separate Linear DOCS issue
• Intend to cherry-pick into the release branch
• I'm proud of this work!