Nexus-Agents Research Index

Generated: 2026-03-18 (ET) Total Papers: 176 | Techniques: 43 | Topics: 12


Quick Stats

StatusPapersTechniques
Implemented-42
In Progress-0
Planned-0
Not Started-0
Rejected-1

Note: Paper-level status tracking deprecated. Technique status is source of truth.


Topics

TopicPapersTechniquesDescription
Consensus106Multi-agent decision protocols and voting
Routing198Cost-efficient model routing and selection
Memory239Context, long-term memory, and compression
Code Generation216Code generation, repair, and self-improvement
CLI Tools00External CLI integration and protocols
Orchestration2712Multi-agent coordination and workflows
Security22Security analysis, prompt injection defense
Evaluation00Benchmarks, metrics, and testing methodologies (planned)
Safety00AI safety, alignment, and reward hacking (planned)
Planning00Task planning, decomposition, and reasoning chains (planned)
Tool Use00Tool augmentation, function calling, and MCP (planned)
Reasoning00Reasoning, self-reflection, and search strategies (planned)

Priority 1 (P1) Techniques

These techniques are high-impact and align well with the current architecture.

TechniqueTopicKey MetricsIssue
Aegean Consensus Protocolconsensuslatency_reduction: 1.2x-20x, token_reduction: 4.4x, quality_impact: within 2.5% of baseline#119
Task-Type Protocol Selectionconsensusreasoning_improvement: +13.2%, knowledge_improvement: +2.8%#125
Multi-Agent Reflexion (MAR)consensusreasoning_improvement: significant across benchmarks-
IPR Quality-Constrained Routingroutingcost_reduction: 43.9%, latency: sub-150ms#128
A-MEM Agentic Memorymemorysemantic_organization: Automatic attribute extraction and linking, evolution_detection: Refinement, extension, supersession detection#122
TRINITY Thinker/Worker/Verifier Rolesorchestrationbenchmark_accuracy: 86.2% on LiveCodeBench#141
Self-Refine Iterative Loopcode-generationaverage_improvement: 20%#126
Reflexion Verbal Reinforcement Learningcode-generationalfworld_improvement: +22%, hotpotqa_improvement: +20%, humaneval_pass1: 91%#130
STPA MCP Frameworksecurityhazard_coverage: Systematic UCA identification, safety_constraints: Auto-generated from analysis#328
AFlow MCTS Workflow Generationorchestrationworkflow_quality: Improved through search optimization#329
SEW Self-Evolving Workflowsorchestrationimprovement_rate: Continuous through execution feedback#330
ZeroRouter Universal Difficulty Spaceroutingrouting_accuracy: Cross-domain difficulty assessment#338
Context Rot Preventionmemory-#1574
Wave-Based Parallel Executionorchestration--

Priority 2 (P2) Techniques

Medium-impact or requiring moderate changes.

TechniqueTopicKey MetricsIssue
CP-WBFT Byzantine Fault Tolerant Consensusconsensusfault_tolerance: 85.7%#103
Free-MAD Anti-Conformity Scoringconsensusrobustness: enhanced against attacks#152
TOPSIS Multi-Criteria Routingroutingcost_reduction: 31.46%#146
PILOT Budget-Constrained Routingroutingadaptive: handles diverse budget requirements#102
SATER Confidence-Aware Routingroutingcost_reduction: 50%+, latency_reduction: 80%+ cascade#99
Agreement-Based Cascading (ABC)routingcost_optimization: significant#121
Preference-Trained Router (RouteLLM)routingcost_reduction: 2x#148
Mem0 Scalable Long-Term Memorymemorylatency_reduction: 91% lower p95, token_savings: 90%, quality_improvement: 26%#156
MIRIX Six-Type Memory Systemmemoryaccuracy_vs_rag: +35%, storage_reduction: 99.9%, benchmark_accuracy: 85.4%#157
MobiMem Post-Deployment Evolutionmemoryprofile_alignment: 83.1%, retrieval_speed: 280x faster than GraphRAG, task_success_improvement: 50.3%#149
Adaptive Memorymemoryperformance_improvement: Configurable priority scoring#143
Evolving Orchestration Upgradeorchestrationtask_completion_improvement: 15-30%#335
LATTS Adaptive Test-Time Computeorchestrationperformance_parity: 1B model matches 405B#153
Voyager Skill Library Patterncode-generationdiscovery_improvement: 3.3x more unique items, speed_improvement: up to 15.3x faster milestone#150
SICA Self-Improving Agentcode-generationswebench_improvement: 17% -> 53%, file_editing_improvement: 82% -> 94%#151
Constitutional AI Self-Critiquecode-generationscales: without human labelers#147
Higher-Order Voting (OW/ISP)consensuscorrelation_handling: Improved consensus on correlated inputs#333
Forest-of-Thought Multi-Tree Reasoningorchestrationreasoning_quality: Improved through diverse exploration, efficiency: Sparse activation reduces compute#331
Agent-SafetyBench Evaluation Suitesecuritycoverage: Multi-dimensional safety evaluation#332
DAAO VAE Difficulty Estimationroutingrouting_accuracy: Improved task-model alignment#334
Hindsight Belief Memorymemoryreasoning_quality: Improved through belief tracking#336
Scaling Agent Coordination Predictororchestrationallocation_efficiency: Improved agent utilization#337
Failure Lesson Injectionorchestration-#1568
Skill Relevance Matchingorchestration-#1569
Write-Time Memory Deduplicationmemory-#1570

Recently Reviewed Papers

DatePaperTopicPriority
2026-03-09Red-Teaming LLM Multi-Agent Systems via Communication Attacksinter-agent-security-
2026-03-09MASFactory: A Graph-centric Framework for Orchestrating LLM-Based Multi-Agent Systems with Vibe Graphinggraph-orchestration-
2026-03-04Style Over Substance: Evaluation Biases for Large Language Models--
2026-03-04FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performancerouting-
2026-03-04CAMEL: Communicative Agents for “Mind” Exploration of Large Language Model Societyorchestration-

Papers by Topic

Consensus (10 papers)

Routing (19 papers)

Memory (23 papers)

Code Generation (21 papers)

Orchestration (27 papers)

Security (2 papers)


GitHub Issues

IssueFeatureRelated Papers
#119Aegean Consensus Protocolarxiv-2512.20184
#125Task-Type Protocol Selectionarxiv-2502.19130
#103CP-WBFT Byzantine Fault Tolerant Consensusarxiv-2511.10400
#152Free-MAD Anti-Conformity Scoringarxiv-2509.11035
#146TOPSIS Multi-Criteria Routingarxiv-2509.07571
#128IPR Quality-Constrained Routingarxiv-2509.06274, arxiv-2406.18510
#102PILOT Budget-Constrained Routingarxiv-2508.21141
#99SATER Confidence-Aware Routingarxiv-2510.05164
#121Agreement-Based Cascading (ABC)arxiv-2410.10347
#148Preference-Trained Router (RouteLLM)arxiv-2406.18665
#156Mem0 Scalable Long-Term Memoryarxiv-2504.19413
#157MIRIX Six-Type Memory Systemarxiv-2507.07957
#149MobiMem Post-Deployment Evolutionarxiv-2512.15784
#142Graph-Based Memoryarxiv-2504.19413
#143Adaptive Memoryarxiv-2310.08560
#122A-MEM Agentic Memoryarxiv-2502.12110
#141TRINITY Thinker/Worker/Verifier Rolesarxiv-2512.04695
#154RL-Trained Orchestratorarxiv-2505.19591
#335Evolving Orchestration Upgradearxiv-2505.19591
#153LATTS Adaptive Test-Time Computearxiv-2509.20368
#126Self-Refine Iterative Looparxiv-2303.17651
#130Reflexion Verbal Reinforcement Learningarxiv-2303.11366
#150Voyager Skill Library Patternarxiv-2305.16291
#151SICA Self-Improving Agentarxiv-2504.15228
#147Constitutional AI Self-Critiquearxiv-2212.08073
#131Self-Debug Code Repairarxiv-2304.05128
#328STPA MCP Frameworkarxiv-2601.08012
#329AFlow MCTS Workflow Generationarxiv-2410.10762
#330SEW Self-Evolving Workflowsarxiv-2505.18646
#333Higher-Order Voting (OW/ISP)arxiv-2510.01499
#331Forest-of-Thought Multi-Tree Reasoningarxiv-2412.09078
#332Agent-SafetyBench Evaluation Suitearxiv-2412.14470
#334DAAO VAE Difficulty Estimationarxiv-2509.11079
#336Hindsight Belief Memoryarxiv-2512.12818
#337Scaling Agent Coordination Predictorarxiv-2512.08296
#338ZeroRouter Universal Difficulty Spacezerorouter-tbd
#1568Failure Lesson Injection
#1569Skill Relevance Matching
#1574Context Rot Prevention
#1570Write-Time Memory Deduplicationarxiv-2601.02553

Search Tags

#adaptive #agent-specialization #agentic-memory #agreement #anti-conformity #attribute-extraction #automatic-discovery #bayesian #belief-state #benchmarking #budget-constraint #byzantine #cascade #code-repair #comprehensive #confidence-aware #constitutional #context-injection #context-management #contextual-bandit #coordination #coordinator #correlation-aware #cost-optimization #cross-critique #debate #deduplication #difficulty-estimation #dynamic #dynamic-linking #dynamic-selection #embedding #ensemble #episodic-memory #evolution #evolutionary #executable #execution-feedback #failure-recovery #fault-tolerance #feedback-learning #formal-safety #formal-verification #graph #hazard-analysis #hindsight #inter-agent #iterative #latent-space #lifelong-learning #lightweight #linucb #long-term #mcp #mcts #memory-efficiency #memory-evolution #multi-criteria #multi-tree #multimodal #observability #optimal-weighting #parallel-execution #pareto #policy #post-deployment #prediction #preference-data #principles #priority #production #protocol-selection #pruning #puppeteer #quality-constrained #quorum #reasoning #recency-decay #reflection #reflexion #reinforcement-learning #relational #relevance #risk-assessment #role-based #routing #rubber-duck #safety-evaluation #scalable #scaling #scoring #self-critique #self-debug #self-evolving #self-feedback #self-improvement #semantic #semantic-similarity #shortest-response #skill-library #sparse-activation #stpa #streaming #structured-memory #task-classification #task-routing #test-time #three-module #token-efficiency #tolerance #training-free #transfer-learning #universal-difficulty #vae #verbal-rl #verifier #versioning #wave-coordination #weighted #worker-dispatch #workflow-generation #workflow-optimization #zettelkasten


Registry Files


How to Contribute

See CONTRIBUTING.md for guidelines on adding new research.


Generated from YAML registries. Last updated: 2026-03-18 (ET)