Back to Ecosystem Pulse
ToolReference
awesome-graph-engineering
by ChaoYue0307
Field guide and CC0 dataset for graph-structured multi-agent systems
Python
Updated Aug 31, 2026
Share:
Overview
Curates a field guide and CC0 dataset for graph-structured multi-agent systems, collecting resources, papers, and design patterns. Organizes agentic systems across nine design layers and eleven sections to help engineers reason about roles, topologies, handoffs, state, gates, reliability, and observability. Includes an interactive atlas interactive atlas and a searchable dataset of 561 resources and 250 papers for design-driven multi-agent engineering.
The Value Proposition
As agents become distributed and interdependent, structure and topology matter as much as individual model quality. This repository surfaces design patterns and empirical literature that make agent interactions, delegation, and failure modes visible and actionable. For trust and evaluation, it provides a shared vocabulary and dataset to benchmark interaction patterns, observability choices, and reliability interventions across multi-agent systems.
Ideal For
Researchers and engineers designing, auditing, or benchmarking multi-agent topologies and workflows who need a curated map of patterns and literature.
How It's Used
- Map agent interaction topologies and reason about delegation and failure modes
- Discover papers and design patterns for building observable, reliable multi-agent workflows
- Create datasets or benchmarks for A2A evaluation and agent track record analysis
Topics
agent-evaluationagent-frameworkagent-orchestrationagentic-aiagentic-workflowsaiai-agentsai-engineeringartificial-intelligenceautonomous-agents+10 more
Keywords
graph-engineeringmulti-agent trustagent-evaluationagentic-workflows