A better mechanism to avoid duplicate AI Client Spans

[alexander-alderman-webb]

The current mechanism to prevent duplicate spans for LLM spans is global integration disabling.
This means that users cannot trace an agent library like langchain together with a client library like openai or anthropic.

The task is to allow users to gather telemetry from different levels of abstractions without a crude method like disabling one integration.

[linear]

PY-2091

[IgorGanapolsky]

Agentic systems are turning into always-on factories: one request fans out into model calls, tool calls, policy checks, and storage.

The hard part is rarely tokens/sec — it is control-plane reliability:

1. Fan-out control — hard cap on tools/steps per request before external writes
2. Durable id + idempotency — one write per business object (ledger / preflight)
3. Observable long jobs — progress + cooperative cancel on multi-minute runs
4. Expected vs actual — durable artifact id before marking success
5. Policy before action — gate that can refuse known-bad side effects

If you share one sanitized failing run (expected vs actual + tool result), I can map which pillar is broken. Architecture only.