A behavioral monitoring technique using HTTP, lexical, and timing signals detects guardrail presence with 100% accuracy and distinguishes guardrail blocks from LLM rejections with 98% average F1 on unseen prompts.
hub
N e M o Guardrails: A Toolkit for Controllable and Safe LLM Applications with Programmable Rails
15 Pith papers cite this work, alongside 91 external citations. Polarity classification is still indexing.
15
Pith papers citing it
91
external citations · OpenAlex
hub tools
citation-role summary
background 1
citation-polarity summary
years
2026 15roles
background 1polarities
background 1representative citing papers
MalSkillBench supplies the first sandbox-verified dataset of malicious agent skills and shows that existing detectors achieve high recall on code injection but collapse on prompt injection and agent-control attacks.
MCPHunt benchmark finds 11.5-41.3% policy-violating credential propagation in multi-server MCP agents across five models, reducible up to 97% by prompt mitigations while retaining most utility.
TokenWall mediates persistent-agent security by auditing source–sink token flows with a local small model and selective large-model escalation, cutting CIK-Bench attack success to 12.5% at low benign latency.
Presents a distributionally robust optimization method for sound probabilistic verification of Datalog policies in AI agents that bounds violation risk regardless of predicate correlations.
Cordon is a transactional runtime system that binds tool intents to reversible state, staged effects, and audit metadata to validate composed agent workflows before commit.
Presents MedSci Skills, an open-source toolkit with deterministic integrity gates for verifying LLM-assisted clinical manuscripts against reporting guidelines like STARD, PRISMA, and STROBE.
Introduces GuardZoo benchmark and RouteGuard router-expert system showing monolithic guardrails suffer task interference while specialized routing improves threat detection and generalization.
PRISM detects and stops credential leakage during LLM generation in multi-agent pipelines using per-token risk scores from lexical, structural, and behavioral signals, achieving zero observed leaks and F1 of 0.832 on a 2000-task benchmark.
Symbolic guardrails enforce about 74% of agent security and safety requirements on three benchmarks with mostly simple checks, improving safety without sacrificing utility.
A code-owned harness enforces source, routing, trace, hygiene, and recommendation contracts for enterprise LLM agents; prompt-only fails and bolt-on guardrails over-refuse.
This survey defines execution provenance as a typed graph of agent execution and evidence tracing as its projection onto evidence-support relations, then reviews methods, taxonomy, benchmarks, and challenges for auditable LLM agents.
Neuro-symbolic pipeline using formal logic and semantic embeddings detects hallucinations in LLM medical reports at 83%+ for entities and 72% for fabrications while cutting creation time 30%.
ADR is a three-component detection system for AI agents that combines telemetry sensors, red teaming, and two-tier detection, achieving 97.2% precision in a ten-month Uber deployment and outperforming baselines on the new ADR-Bench.
TWGuard achieves +0.289 F1 improvement and 94.9% false-positive reduction for LLM safety guardrails in the Taiwan linguistic context compared to foundation models and baselines.
citing papers explorer
-
Behind the Refusal: Determining Guardrail Activation via Behavioral Monitoring
A behavioral monitoring technique using HTTP, lexical, and timing signals detects guardrail presence with 100% accuracy and distinguishes guardrail blocks from LLM rejections with 98% average F1 on unseen prompts.
-
MalSkillBench: A Runtime-Verified Benchmark of Malicious Agent Skills
MalSkillBench supplies the first sandbox-verified dataset of malicious agent skills and shows that existing detectors achieve high recall on code injection but collapse on prompt injection and agent-control attacks.
-
MCPHunt: An Evaluation Framework for Cross-Boundary Data Propagation in Multi-Server MCP Agents
MCPHunt benchmark finds 11.5-41.3% policy-violating credential propagation in multi-server MCP agents across five models, reducible up to 97% by prompt mitigations while retaining most utility.
-
Token-Flow Firewall: Semantic Runtime Auditing for Persistent AI Agents
TokenWall mediates persistent-agent security by auditing source–sink token flows with a local small model and selective large-model escalation, cutting CIK-Bench attack success to 12.5% at low benign latency.
-
Efficient and Sound Probabilistic Verification for AI Agents
Presents a distributionally robust optimization method for sound probabilistic verification of Datalog policies in AI agents that bounds violation risk regardless of predicate correlations.
-
Cordon: Semantic Transactions for Tool-Using LLM Agents
Cordon is a transactional runtime system that binds tool intents to reversible state, staged effects, and audit metadata to validate composed agent workflows before commit.
-
Deterministic Integrity Gates for LLM-Assisted Clinical Manuscript Preparation: An Auditable Biomedical Informatics Architecture
Presents MedSci Skills, an open-source toolkit with deterministic integrity gates for verifying LLM-assisted clinical manuscripts against reporting guidelines like STARD, PRISMA, and STROBE.
-
Triaging Threats to Specialized Guardrails
Introduces GuardZoo benchmark and RouteGuard router-expert system showing monolithic guardrails suffer task interference while specialized routing improves threat detection and generalization.
-
PRISM: Generation-Time Detection and Mitigation of Secret Leakage in Multi-Agent LLM Pipelines
PRISM detects and stops credential leakage during LLM generation in multi-agent pipelines using per-token risk scores from lexical, structural, and behavioral signals, achieving zero observed leaks and F1 of 0.832 on a 2000-task benchmark.
-
Don't Make Models Guess Security and Safety: Symbolic Guardrails for Domain-Specific AI Agents
Symbolic guardrails enforce about 74% of agent security and safety requirements on three benchmarks with mostly simple checks, improving safety without sacrificing utility.
-
From Prompts to Contracts: Harness Engineering for Auditable Enterprise LLM Agents
A code-owned harness enforces source, routing, trace, hygiene, and recommendation contracts for enterprise LLM agents; prompt-only fails and bolt-on guardrails over-refuse.
-
From Agent Traces to Trust: A Survey of Evidence Tracing and Execution Provenance in LLM Agents
This survey defines execution provenance as a typed graph of agent execution and evidence tracing as its projection onto evidence-support relations, then reviews methods, taxonomy, benchmarks, and challenges for auditable LLM agents.
-
Neuro-Symbolic Verification of LLM Outputs for Data-Sensitive Domains (extended preprint)
Neuro-symbolic pipeline using formal logic and semantic embeddings detects hallucinations in LLM medical reports at 83%+ for entities and 72% for fabrications while cutting creation time 30%.
-
ADR: An Agentic Detection System for Enterprise Agentic AI Security
ADR is a three-component detection system for AI agents that combines telemetry sensors, red teaming, and two-tier detection, achieving 97.2% precision in a ten-month Uber deployment and outperforming baselines on the new ADR-Bench.
-
TWGuard: A Case Study of LLM Safety Guardrails for Localized Linguistic Contexts
TWGuard achieves +0.289 F1 improvement and 94.9% false-positive reduction for LLM safety guardrails in the Taiwan linguistic context compared to foundation models and baselines.