Sentinel Agents for Secure and Trustworthy Agentic AI in Multi-Agent Systems
cs.AI, cs.MA
Submitted: 2025-09-18
Updated: 2026-11-05
Comments: 23 pages, 11 figures. v2: the preliminary evaluation of v1 is replaced by a new evaluation on a balanced 1,000-message corpus with ablation, local LLM benchmark, aggregation and calibration analyses; paper restructured; references updated
Code: https://github.com/open-voice-interoperability/openfloor-docs
License: http://creativecommons.org/licenses/by/4.0/
The gist: Multi-agent systems (MAS) built on large language models (LLMs) increasingly share a common conversational space…
Terminology
Abstract
Multi-agent systems (MAS) built on large language models (LLMs) increasingly share a common conversational space, in which a single adversarial or erroneous message can propagate from one agent to many. This paper proposes Sentinel Agents, a dedicated security layer that monitors inter-agent communication with rule-based checks, behavioral analysis and LLM-based semantic inspection, and reports to a Coordinator Agent that enforces policy, for example by quarantining misbehaving agents. We describe the architecture, its deployment patterns and monitoring layers, and we evaluate a continuous-listener prototype on a balanced corpus of 1,000 synthetic messages (500 attacks across five families and 500 benign messages) with two local LLMs. LLM inspection with Gemma 4 12B reached F1 0.86 at a false-positive rate of 0.11, against 0.38 for rules alone. Combining components by logical OR did not improve on it, while a cascade in which only high-precision rules act on their own matched it. Calibrating the thresholds of the more conservative Qwen3 8B on a development split raised its held-out F1 from 0.56 to 0.80, and on multi-turn conversations the behavioral check detected patterns the rules missed but could not tell legitimate repetition from probing. The value of combining detectors therefore depends on the model, on its calibration and on the aggregation rule, and the results yield practical guidance for deploying Sentinel Agents.
Sources
- Hallucination Mitigation using Agentic AI Natural Language-Based Frameworks
- Conversational AI Multi-Agent Interoperability, Universal Open APIs for Agentic Natural Language Multimodal Communications
- AI Multi-Agent Interoperability Extension for Managing Multiparty Conversations
- Prompt Injection Detection and Mitigation via AI Multi-Agent NLP Frameworks
- Modeling Response Consistency in Multi-Agent LLM Systems: A Comparative Analysis of Shared and Separate Context Approaches
- Multi-Trigger Poisoning Amplifies Backdoor Vulnerabilities in LLMs
- Security of Internet of Agents: Attacks and Countermeasures
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents