feat(moa): add reference_max_tokens to cap advisor output and cut turn latency (#56756)

MoA per-turn latency is dominated by advisor GENERATION: turn wall time
correlates ~0.88 with output tokens and ~-0.03 with input tokens (measured over
52 turns). Each turn waits for the slowest advisor to finish writing, and
advisors were uncapped — writing multi-thousand-token essays the aggregator
only needs the gist of.

Add an opt-in per-preset reference_max_tokens knob (mirrors reference_temperature)
that caps ADVISOR output only; the acting aggregator is never capped. Default
None = uncapped, so existing presets are byte-for-byte unchanged (no regression).
Wired through both MoA execution paths (MoAChatCompletions.create and
aggregate_moa_context).

E2E: same task, closed preset uncapped vs reference_max_tokens=600 -> 59s to 33s
(~44% faster), final answer identical/correct.

- hermes_cli/moa_config.py: _coerce_int_or_none helper + reference_max_tokens
  in _normalize_preset/_default_preset/flattened view
- agent/moa_loop.py: read preset.reference_max_tokens, pass to reference fan-out
- agent/conversation_loop.py: pass reference_max_tokens on the per-turn path
- tests + docs

agent/conversation_loop.py

@@ -856,6 +856,7 @@ def run_conversation(
                     aggregator=moa_config.get("aggregator") or {},
                     temperature=float(moa_config.get("reference_temperature", 0.6) or 0.6),
                     aggregator_temperature=float(moa_config.get("aggregator_temperature", 0.4) or 0.4),
+                    max_tokens=moa_config.get("reference_max_tokens"),
                 )
                 if _moa_context:
                     for _msg in reversed(api_messages):