Skip to content
Preprint

Context-to-Execution Integrity for LLM Agents

Jul 2026 · 4 citations · ⚡ 1 influential · 53 references
Computer Science

TL;DR

Across the evaluated sinks, CXI admits execution only when field, effect, and invocation authority bind to the same action manifest, and across the evaluated sinks, CXI admits execution only when field, effect, and invocation authority bind to the same action manifest.

Abstract

Language-model agents read attacker-writable context to solve tasks. Tool execution needs a separate authority check for protected sink fields, sink-interpreted payloads, and the invocation event. Context-to-Execution Integrity (CXI) is an execution-boundary system for this setting. Policies mark protected sink fields, typed releases carry narrow validated values from writable context to specific destinations, opaque data slots keep evidence as data, and a deterministic gate admits a call only after field authority, exact-effect authorization, and invocation authority all bind to the same action manifest. We evaluate CXI on open-weight field-projection runs, AgentDojo live episodes, a code-agent exact-effect benchmark, manifest-bound ledger faults, proposal-pressure controls, and hosted/API compatibility traces. AgentDojo covers 720 live episodes and 1,739 LLM calls; the code-agent benchmark covers 400 repository episodes with exact-effect authorization and lease-bound execution, yielding 231 safe task completions and zero observed field, effect, or invocation escapes. The accounting reports parser outcomes, authorization outcomes, and task-quality outcomes together with the admission-integrity result. Across the evaluated sinks, CXI admits execution only when field, effect, and invocation authority bind to the same action manifest.

View source

Similar papers

Preprint Aug 2026

AgentFlow: A Flow-Centric Policy Language and Framework for Securing LLM Agent Systems

AgentFlow, a flow-centric policy language and runtime enforcement model for specifying where data may travel in agent systems, is presented and results are preliminary and scoped to the modeled policy-visible agent behaviors and evaluated benchmarks.

B. Shivakumar, Swarn Priya, Peng Gao · 0 citations
Preprint Jul 2026

FAVA: Formal Authorization for Verified Agents with Evidence-Backed Permission Graphs

This work presents FAVA (Formal Authorization for Verified Agents), a permission-carrying authorization framework for agent execution that utilizes an LLM-guided Permission Intermediate Representation to translate ambiguous natural-language tasks into structured constraints.

Yifan Zhang, Xinkui Zhao, Sai-Da Liu et al. · 3 citations
Preprint Jul 2026

ContainmentBench: Trace-Based Evaluation of Post-Exposure Containment in Tool-Using LLM Agents

ContainmentBench, a sandboxed benchmark comprising a 504-scenario specification dataset, a shared rollout-trace schema, and stage-scoped metrics for endpoint violations, logged propagation, and explicitly authorized taint-exposed proposals that commit, is introduced.

Wenhao Lan, Shan Li, Meiqi Wu et al. · 0 citations
Preprint Aug 2026

SPA: Securing Persistent LLM Agents Across Queries with Plan-First Information-Flow Control

Large language model (LLM) agents increasingly operate over untrusted webpages, documents, tools, and persistent states while exercising authority over security-sensitive resources. Existing defenses typically protect either planning or individual tool interactions, but persistent agents face a broader threat: attacker-controlled data can alter control flow, enter security-sensitive tool arguments, or compromise later queries. We present SPA, a plan-first architecture that secures planning, execution, and cross-query state reuse. SPA invokes the planner once per query to generate a complete executable plan in a declarative domain-specific language, then applies dual-lattice information-flow control to track confidentiality and integrity across explicit data flows and control dependencies. To support persistence without re-exposing untrusted payloads to the planner, SPA stores execution results as labeled artifacts and reveals only semantic metadata during later planning. We evaluate SPA on AgentDojo and AgentDojo-MQ, which is our multi-query extension for measuring secure state reuse and delayed attacks. Under the'tool_knowledge'attack, SPA with information-flow control reduces attack success to zero on AgentDojo and 0.2% on AgentDojo-MQ. Our results show that plan-first execution combined with label-preserving persistence can substantially strengthen persistent LLM agents, while revealing an important security-utility tradeoff introduced by strict integrity enforcement.

Dylan Girrens, Guangshen Wang · 0 citations
Preprint Jul 2026

Agentao: A Policy-Governed Runtime Harness for Embeddable Tool-Using LLM Agents

Agentao is presented, a governed local-first runtime for tool-using LLM agents that separates model-generated action proposals from host-authorized execution through a layered architecture consisting of host-facing surfaces, a host contract, a runtime core, a permission-mediated tool system, and supporting subsystems for memory, replay, plugins, skills, sub-agents, and protocol integration.

Bo Jin, Q. Jiao, Xin Tong · 1 citation