GNAP: trace git-native agent coordination events alongside RagaAI's agentic tracing

Proposal: RagaAI Catalyst integration for GNAP multi-agent coordination tracing

RagaAI Catalyst provides comprehensive LLM project management — trace management, agentic tracing, evaluation, and guardrails. Your agentic tracing feature is particularly relevant: as multi-agent systems grow more complex, coordination events between agents need to be traced alongside LLM calls.

GNAP (Git-Native Agent Protocol) coordinates agents via git: tasks move through board/todo/ → board/doing/ → board/done/. Each transition is a natural trace event that complements RagaAI's existing span/trace model.

GNAP + RagaAI Catalyst integration:

from ragaai_catalyst import RagaAICatalyst
from ragaai_catalyst.tracers import Tracer

catalyst = RagaAICatalyst(access_key=..., secret_key=..., base_url=...)
tracer = Tracer(project_name="multi-agent-system")

with tracer.trace():
    # GNAP coordination events are traced alongside LLM calls
    task = gnap_claim_task("board/todo/analyze-customer-data.md")
    # tracer captures: gnap.task.claim event with metadata
    
    result = llm_agent.process(task)  # LLM calls traced by RagaAI
    
    gnap_complete_task(task, result)
    # tracer captures: gnap.task.complete event

RagaAI's trace management would then show the full picture: coordination latency (time in each GNAP state) + LLM call metrics side by side. This enables root cause analysis like "is slowness due to coordination bottleneck or LLM latency?"

A GNAP integration in RagaAI Catalyst would make it the most complete observability platform for multi-agent systems — from coordination to LLM to evaluation.

Spec: https://github.com/farol-team/gnap