AI Agent Security Benchmarks

Large language model (LLM) agents are starting to take on routine work in high-performance computing (HPC), including monitoring Slurm jobs, diagnosing failed builds, inspecting simulation output, and coordinating scientific workflows. To do this work, an agent commonly acts under its user's credentials and inherits the user's access to files and the scheduler. This arrangement creates a failure mode that ordinary account-level controls do not capture. Adversarial instructions in a log, tool description, shared file, or peer-agent message may redirect the agent beyond the task the user assigned, even though every resulting command is authenticated and permitted for that account. We refer to this as the hijacked authorized agent problem. Existing agent-security studies explain relevant mechanisms, such as indirect prompt injection and tool misuse, but generally evaluate them in web, enterprise, or personal-assistant settings. HPC security, by contrast, has mature controls for identity and isolation but does not ordinarily represent the intent of a particular task. This paper defines the threat model in the HPC setting, identifies attack surfaces created by schedulers, shared storage, multi-project accounts, and scientific workflows, and examines where current controls fall short. It concludes with a research agenda and a plan for an empirical benchmark, TaskBound.

LLM-based agents process external content, exposing them to prompt injection and multi-turn manipulation. Most safety benchmarks evaluate defenders against fixed attack pools collected before evaluation, single-turn or multi-turn. We present a 21-scenario benchmark for \emph{adaptive multi-round attacks against memoryless LLM defenders}: an autonomous LLM attacker observes prior defender responses and pivots across rounds, while each defender response is evaluated as a fresh interaction. Holding the 21 scenarios, attackers, defenders, and structured-output scoring fixed, restricting scoring to the first attacker turn yields $0$-$1\%$ attack success rate (ASR); allowing 15 rounds of adaptive attack yields $5.4$-$14.0\%$. Pooling three frontier attacker LLMs uncovers $1.4$-$2.2\times$ as many unique successful attacks as the best single attacker, and the generated attacks have low cosine similarity ($0.02$-$0.14$) to attacks in existing benchmarks. Claude Opus 4.6 and GPT-5.4 are tied in aggregate ($5.4\%$ each; overlapping $95\%$ CIs), but their weaknesses differ sharply: on one scenario Opus reaches $60\%$ ASR ($95\%$ CI $36$--$80\%$) while GPT-5.4 and Gemini each stay at $7\%$ (CI $1$-$30\%$; the gap is preserved in a higher-$N$ replication). $13$ of $21$ scenarios distinguish at least one defender pair, yet rankings disagree across scenarios (Kendall's $W = 0.19$). We release the benchmark -- 21 evaluation scenarios, 10 public development scenarios, the orchestrator, baseline harnesses, and a multi-attacker CLI -- plus 945 transcripts from the 3$\times$3 frontier matrix, an attack-replay dataset, and 18{,}422 gpt-oss-20b battles from an open competition's final scoring rounds.

Persistent external memory enhances agent continuity but introduces persistent security vulnerabilities: adversarial content can be injected via standard interaction channels, retained across turns, and later distort downstream behavior. To address this challenge, we propose MemPoison, a comprehensive benchmark and analysis framework featuring 1227 hand-validated cases across four attack types, three injection channels, and three representative memory substrates, evaluated on seven open-weight and three closed-weight model families. We introduce a three-tier taxonomy: (L1) direct single-record corruption, (L2) compositional multi-record corruption and (L3) context-triggered dormant corruption. Our evaluations reveal a distinct defense frontier: while baseline write-time defenses, such as consistency checks, substantially suppress direct L1 attacks, they fail to reliably suppress L2 and L3 attacks. Through mechanistic influence decomposition (MID), we demonstrate structural blind spots in write-time defenses, which admit seemingly benign records that later become harmful through joint retrieval composition or trigger-conditioned activation. Our findings advocate for shifting from static filtering to adaptive, context-sensitive memory defense strategies.

Clinical AI models can expose patients to harm when adversarial vulnerabilities go undetected, yet formal security auditing requires statistical expertise, specialized tools, and significant time. We present an open evaluation task, built on METR Task Standard v0.3.0, that tests whether frontier AI agents can autonomously implement a structured clinical AI security audit. Given a pre-trained clinical prediction model, a patient dataset, and written instructions, each agent must implement four attacks from pseudocode, compute a Security Posture Score covering FGSM robustness, membership inference resistance, expected calibration error, and boundary attack resistance, and write a structured JSON report in a Docker container using only a bash interface and no scaffolding code. Six variants span the Wisconsin Diagnostic Breast Cancer and MIMIC-IV ICU mortality datasets across three model architectures with increasing defense strength, with reference scores from 55.60 to 90.41. We ran 54 evaluations across three frontier models, with three runs per variant. Claude Sonnet 4.6 and GPT-4.1 completed all 18 runs and received perfect evaluator scores. GPT-4o completed 61 percent of runs and used about five times the per-run token count of Claude, although provider tokenization differs. Total API costs were 8 US dollars for GPT-4.1, 12 US dollars for Claude Sonnet 4.6, and 27 US dollars for GPT-4o. GPT-4o failures involved premature session termination, an aggregation error, and an empty submission file. The task, scoring infrastructure, and Wisconsin Breast Cancer assets are publicly released; MIMIC-IV variants require separate PhysioNet access.

Recent autonomous penetration testing papers report high benchmark scores while adding multi-component security harnesses around frontier LLMs. Because these systems often change both architecture and backbone model, it is difficult to tell how much performance comes from the harness rather than from the underlying model. This paper presents a controlled study on the 104-task XBOW benchmark using default coding CLI agents as plain-agent baselines. We first run Codex, OpenCode, and Pi with the same GPT-5 model, budget, target interface, and scoring rule. This phase identifies the strongest same-model baseline and tests whether security-specific prompt variants improve its observed score. We then compare the default Codex scaffold with published MAPTA and PentestGPT V2 results under the closest available model matches. Finally, we repeat the plain-agent experiment with GPT-5.2 and GPT-5.5 to measure model scaling inside the same scaffold. The results show a mixed but practical picture. Specialised harnesses can add measurable benchmark lift and may improve cost efficiency, but plain coding agents already solve a large share of the benchmark; repeated plain-agent runs can match or exceed some published architecture scores in union coverage, and newer models substantially improve the same scaffold. Future evaluations should report model-matched plain-agent baselines before attributing benchmark gains to architecture design alone.

We present nsfaguard, a guardrail framework for securing agentic AI systems against operational threats, such as prompt injection, sensitive information extraction, malicious code requests, dangerous tool misuse, and resource exhaustion. We first introduce the NSFA taxonomy, which organizes 185 risk variants into a CIA-triad-grounded hierarchy and is cross-validated against three well-established OWASP guidelines. Based on this taxonomy, we construct a benchmark suite spanning 133 languages, comprising over 93K purpose-built samples targeting both user queries and agent responses, along with 3,435 cross-source samples adapted from five public agent-security datasets. To detect these operational threats in practice, we develop a dual-mode approach combining SFT-based generative reasoning for interpretable offline auditing with discriminative classification heads on the frozen backbone, enabling real-time detection at approximately 50,ms. We release four models with 0.8B, 2B, 4B, and 9B parameters, all achieving $\geq$94% F1 on purpose-built benchmarks and surpassing the strongest competing guardrails by 6 to 12 absolute points. On cross-source evaluation, the 9B model attains 91.29% F1 with a more balanced precision--recall trade-off. Moreover, ablation experiments show that classification heads can equip a guardrail with risk detection capabilities beyond its original scope and achieve state-of-the-art performance. These results demonstrate the extensibility of the approach and its generality as a plug-in enhancement.

Tool-using large language model (LLM) agents are attractive for network operations, but tickets, alerts, logs, runbooks, and ChatOps messages can carry indirect prompt injections. We present NetInjectBench, a 130-scenario benchmark that separates untrusted artifact text, trusted policy metadata, and evaluation labels for network-operation tool use. The sample contains 40 benign, 40 weak-attack, 40 strong-attack, and 10 approved high-impact change scenarios; each is evaluated with Qwen2.5-7B, Llama3.1-8B, and Mistral-7B. Across 240 attack instances, naive execution reached an 82.50% unsafe tool-action rate. Prompt-only safety, Self-Reminder, Spotlighting, and a Two-Pass LLM Judge reduced this rate to 25.63%, 21.67%, 18.33%, and 10.00%, respectively. Static allowlisting reached 5.00% but blocked all approved changes, yielding 0.00% usefulness and 100.00% overblocking on approved cases. Under the stated metadata-integrity assumption, the metadata-aware policy gate produced 0/240 unsafe attack actions, with a 95% Wilson upper bound of 1.58%, while preserving 99.17% attack-scenario usefulness and 100.00% approved-change usefulness. The findings show that network-operation agents need execution-time authorization boundaries alongside prompt-level instruction hygiene.

LLM agents can commit durable effects from authority evidence that was valid earlier in execution: a DOM snapshot, approval epoch, version witness, branch token, or worker result. We study the commit boundary at which earlier authority evidence no longer authorizes a durable effect. We call this property commit-time authorization: a durable effect is authorized only if the witness that licensed its derived state remains fresh, causally prior, bound to the same effect, and eligible at commit time. We build a controlled-invalidation suite spanning browser, tool/API, and multi-agent workflows. The suite preserves the user goal and payload shape while invalidating the authority relation before durability. In the primary 54-task matrix, endpoint success remains high: 262/270 runs reach the visible result. Only 55/270 are authorized completions; among the 216 invalidating rows, 207 commit after the authorizing path has failed. All 54 clean controls remain authorized, and a separate 54-run authority-preserving check produces no unauthorized commits. We then evaluate mitigation families. Prompt caution and single-condition checks are insufficient because different hazards break different boundary conditions. Defenses work when they refresh, rebind, replan, or refuse at the durability boundary. CommitGuard, a fail-closed boundary monitor, blocks stale durable-effect attempts on protected commit surfaces when runtimes emit witness, dependency, binding, and eligibility signals. The result is a reporting and runtime-design lesson: endpoint success is a utility metric; authorized commit is a security property.

Agentic red-teaming benchmarks report whether an injected agent was compromised as a single bit: the attack succeeded, or it did not. We argue that this binary attack-success rate discards the information a defender most needs, namely how harmful the resulting action was. We introduce an action-graded harm rubric that scores an agent's tool-call trajectory on a seven-level ordinal scale (L0 to L6) according to whether the executed action was reversible, whether it crossed scope to reach another party, and whether it expanded privilege. We compute the scale two ways: a deterministic oracle that reads the trajectory and the attacker's stated goal, and a panel of three frontier language-model judges that read a tag-free account of the same trajectory. Across four victim models and two defenses on the AgentDojo workspace suite, severity grading exposes three cases the binary metric hides, including a defense that reports a zero attack-success rate while still permitting an externally visible cross-scope leak through an unfiltered tool. The judge panel reproduces the oracle with high ordinal agreement (Krippendorff's alpha = 0.91) but shares systematic blind spots that we characterize, most notably a failure to recognize escalation chains. Unlike prior work that provides harm taxonomies, harmful-task completion tests, execution-level safety benchmarks, or severity-aware simulation, our contribution is a reusable, trace-grounded severity instrument applied to the actual actions recorded in existing red-team logs. All code, prompts, and per-episode logs are released.

As LLM agents evolve from single-user assistants into shared organizational infrastructure, new privacy risks emerge: inappropriate information may not only be exposed through outputs for external recipients, but also internally across users through inter-agent messages, shared memory and agents. These data spillage risks are not captured by existing privacy benchmarks grounded in contextual integrity (CI) as they focus primarily on either single-user settings or interactions between independently owned agents. We introducePiSAs (Privacy in Shared Agentic systems), a benchmark for assessing unintentional leaks with dual CI annotations: whether an information is appropriate for the task, and which users may legitimately access it. This enables direct measurement of cross-user spillage across agentic system components and interfaces, such as outputs, inter-agent communication, and memory. PiSAsis system-agnostic and supports evaluation across different agent topologies and memory regimes. We find that, although system design improves CI compliance, results are bottlenecked by incorrect LLM judgment calls: even state-of-the-art models fail to reliably filter inappropriate content or restrict transmission to authorized users. Our findings underscore the need for privacy-preserving strategies, beyond those studied in this work.

Persistent personal agents combine long-term memory with access to users' external environments, enabling personalized foreground assistance and proactive background execution. This integration also creates a new path to compromise: untrusted external content can be silently written into persistent memory and later reused as trusted state. We study this threat as stealth memory injection, in which a remote black-box adversary delivers a single email payload that must induce the agent to write poisoned memory, stay hidden in the agent's response to the user, and affect future behavior. We introduce WhisperBench, a 108-case benchmark spanning five risk categories and both fact and preference poisoning. Built on a real IMAP/SMTP workflow and an authentic email agent skill, it enables full-cycle evaluation of stealth memory injection attacks. To enable this black-box attack under single-email delivery and without runtime feedback, we propose MemGhost, a one-shot payload generation framework. MemGhost uses an environment proxy to emulate persistent-agent execution and an objective proxy to convert memory adoption and conversational stealth into dense rubric-based rewards, then trains the attacker policy with supervised fine-tuning and reinforcement learning. Across 56 held-out test cases, MemGhost achieves 87.5% end-to-end success on OpenClaw with GPT-5.4 and 71.4% on Claude Code SDK with Sonnet 4.6. It also transfers across personal-agent architectures (NanoClaw and Hermes Agent) and memory backends (filesystem and vector-based Mem0), and remains effective against input-level, model-level, and system-level defenses. These results suggest that persistent memory can turn ordinary external processing into a practical pathway for long-term agent compromise.

Recent tools such as OpenClaw have extended the capabilities of LLM-based agents from simple dialog-based systems to fully autonomous agents. These systems allow personalization of the agent through modifiable internal files and the installation of skills. While this enables deployment in a wide range of settings and the automation of diverse tasks, greater capability and autonomy increases the risk of malicious actions being executed unintentionally. In this work, we explore the interplay between agent configuration and the risk of executing dangerous actions without explicit instruction. To this end, we propose CONfiguration Tree-search for Red-teaming Agents (CONTRA), an LLM-assisted tree-search algorithm that discovers agent configurations resulting in the execution of malicious actions. CONTRA works by reasoning about benign yet dangerous configurations and evaluating them in a simulated environment. We construct a dataset of the 473 most popular skills from a public repository, along with 2-5 corresponding malicious target actions per skill. In a large-scale analysis, we find that 75.1% of skills have at least one configuration resulting in the execution of a malicious action, most of which have not been detected as containing malicious content by existing scans. Overall, CONTRA successfully identifies a configuration leading to the execution of the target action in 39.2% of all tested cases. Our findings demonstrate that current agents provide insufficient safety with respect to personalization.

As AI coding agents become more autonomous, they increasingly ship code iteratively, with the codebase persisting across sessions. This persistence creates a new attack surface: a misaligned or prompt-injected agent can distribute attacks across pull requests (PRs) and time its payload for the PR with the best natural cover. To study the resulting dynamics, we introduce Iterative VibeCoding, a setting for AI control, the study of safely deploying capable but potentially untrusted AI. In Iterative VibeCoding, a coding agent builds software over a sequence of PRs in a persistent codebase while pursuing a covert side task. Our benchmark includes two task families: CLI tools and Flask web services, across 20 total task variations. We use Claude Sonnet 4.5 as the attack agent and GPT-4o as the monitor. We compare gradual attacks, which distribute the side task across PRs, against non-gradual attacks concentrated in a single PR. No single monitor is robust to both: which strategy evades best (success while evading the monitor) depends on the monitor type, so a defender cannot close off both gradual and non-gradual attacks with any one monitor. High evasion (>= 65%) generalizes across model attack agent backends (Sonnet 4.5, Gemini 3.1 Pro, Kimi K2.5), confirming this is a property of the persistent-state attack surface rather than a single model's capability. Evasion also remains high across state-of-the-art monitor models and the gap between gradual and non-gradual evasion widens for more capable models. We introduce a stateful link-tracker monitor that tracks suspicious buildup across PRs. On both task families, it detects gradual attacks substantially better than diff monitors that merely see more accumulated history. Combining this stronger monitor with trajectory monitors in a four-monitor ensemble reduces gradual-attack evasion from 93% under the weakest standard diff monitor to 47%.

Claw-like AI agents (e.g., OpenClaw) are always-on processes with persistent access to credentials, files, tools, and external services. They take on system-level responsibilities -- installing packages, maintaining state, scheduling subtasks, and mediating I/O -- making security failures far more severe than in other agents. Yet existing benchmarks focus on model responses and tool calls, leaving cross-component failure modes largely unmeasured. We adopt a computer-system analogy: treating a Claw-like agent as an agentic computer system whose gateway runtime plays an OS-like mediation role, whose Skills resemble user-installed applications, and whose Plugins resemble loadable extensions with runtime privileges. Each component has a classical counterpart whose protection mechanisms -- refined over decades of cybersecurity research -- are absent on the agent side. From this perspective, we develop SafeClawArena, a benchmark of 406 adversarial tasks across four attack surfaces (Skill Supply-Chain Integrity, Persistent State Exploitation, Cross-Boundary Data Flow, and Indirect Prompt Injection), executed in containerized replicas of real agent platforms with canary-marked credentials and evaluated via automated taint tracking across nine output channels. We evaluate three platforms (OpenClaw, NemoClaw, SeClaw) and five frontier LLMs. The highest attack success rate reaches 70%; malicious Plugins succeed in 100% of cases regardless of the LLM. SeClaw cuts GPT-5.4's attack success rate from 70% to 22%, partly through utility-security tradeoffs rather than active defenses, while Claude-Opus-4.6 already sits near a 22% floor on every platform. These results expose the inadequacy of current defenses and suggest directions for future hardening. Code and data: https://github.com/sunblaze-ucb/SafeClawArena.

Model Context Protocol (MCP)-style ecosystems give language-model applications a practical connection layer for tools, resources, prompts, and transports. As agents move from connection to execution, security decisions often remain split across clients, servers, prompts, approval dialogs, OAuth deployments, and logs. This paper asks whether a runtime can make execution-layer invariants explicit and testable while preserving MCP-like workflows. We define eight invariants: metadata non-authority, grant-backed approval, canonical resources, principal binding, scoped capability invocation, source-and-target data-flow authorization, deny-path audit, and explicit protocol state. We implement these invariants in HCP, a Handle-Capability Protocol reference runtime for MCP-style agent execution that represents calls through principals, resources, grants, capabilities, handles, policy decisions, data-pipe checks, and audit entries. We evaluate HCP against two MCP-like baselines: a naive connection-layer runtime and a practice-informed connection-layer mitigation baseline with metadata linting, session checks, and per-call approvals. Across 10 benchmark cases, the naive baseline permits all modeled attacks, the mitigation baseline permits 6 of 10, and HCP blocks all 10 while preserving audit evidence. Ablations identify which runtime components block attacks and preserve forensic evidence. A local in-memory microbenchmark reports sub-millisecond mean latencies for measured policy, invocation, peek, and pipe operations. A bounded GitHub README-screening sample provides ecosystem signals, not vulnerability findings. The results support a narrow claim: MCP-style agent systems need an execution-control layer in addition to connection-layer conventions.

Multimodal agentic retrieval-augmented generation (RAG) systems expand the attack surface beyond prompt injection to include text poisoning, image injection, direct-query attacks, and orchestrator-level tool manipulation. Existing red-teaming approaches are typically surface-specific and often recycle known attack templates; on text-poisoning benchmarks we measure 73-84% exact duplication. We present MIRROR, a unified cross-surface framework that performs memory-guided Monte Carlo tree search while conditioning candidate generation on retrieved context under an explicit novelty constraint. A deterministic Novelty Gate rejects any candidate matching the retrieval set under normalized comparison, allowing retrieval to inform search priors without enabling prompt copying. Across four attack surfaces on a multimodal agentic RAG target, MIRROR attains 76% ASR on image poisoning compared with 52% for baselines, 97% ASR on orchestrator attacks at half the query cost, and the lowest cross-surface variance (coefficient of variation 0.47). In contrast, specialized baselines collapse across surfaces: suffix optimization reaches 79% ASR on text poisoning but 1% on direct queries. We release ART-SafeBench with 41,815 in-package records and runtime adapters yielding 41,991+ total records across four surfaces.

We present CyberChainBench, a benchmark for evaluating LLM-based agents on smart contract security across three complementary tasks: vulnerability detection, exploit generation, and patch synthesis. Built from 541 real-world exploit incidents from DeFiHackLabs spanning 9 EVM chains, the benchmark provides end-to-end on-chain evaluation where agents interact with historical blockchain state through isolated evaluation environments orchestrated by Harbor, using tools to read code, trace transactions, and validate exploits on mainnet forks. Each case is anchored to a specific block and includes structured ground truth covering vulnerability type, localization, and attacker profit. Exploits are graded by economic impact on historical forks; patches are validated by replaying historical attacks and legitimate transactions as fail-to-pass test oracles on a proxy-upgradeable subset. We define a five-type vulnerability taxonomy and evaluate multiple agent--model configurations. Results reveal a clear difficulty gradient: the best configuration scores 37.5% on detection, 43.7% on exploitation, but only 23.4% on patching, with the top agent (Codex with GPT-5.5) realizing $57.4M in total exploit profit across the 200-case exploit set at a cost of $2.39 per case.

To better assist users with completing challenging tasks, AI agents mediate communications, access data, and interact with different APIs. Many employers (and even nation-states) already provide their users with this technology. However, widespread adoption of AI agents creates a new risk to abuse access to user data for another goal: surveilling users. These users might not even have the ability or permission to control the actions and data accesses of the surveilling agents. We introduce and formalize the problem of agentic surveillance: the ability of an AI agent to analyze available information, craft a report, and send it out using available tools. To evaluate surveillance capabilities across different models, we create SurveilBench, a dataset of various reporting scenarios focusing on three domains: corporate, education, and police. We find that some models exhibit emergent (i.e., unprompted) tendencies to help surveillance, but they also report the attempts to surveil users to the government. Finally, we repurpose prompt injections for evading surveillance and develop three evasion techniques that hide from, deceive, or induce over-escalation in surveillance agents. We conclude that agentic surveillance can already be easily implemented and, therefore, call for a comprehensive technical, ethical, and legislative framework to protect users.

Large Language Models (LLMs) have shown promise for automated penetration testing, yet existing end-to-end black-box evaluations are highly susceptible to error cascading: failures in early reconnaissance can mask an agent's actual ability to exploit vulnerabilities. To more accurately characterize these capabilities, we propose a two-stage decoupled evaluation framework that separates exploit execution from reconnaissance. Using ground-truth injection and knowledge-driven ablation across 70 high-fidelity web vulnerability testbeds, our framework isolates exploitation performance from reconnaissance noise. We empirically evaluate five open-source penetration-testing agents, covering multiagent, monolithic, and graph-driven architectures, on a strictly aligned subset of 50 representative vulnerabilities. The results reveal a substantial capability gap. With accurate vulnerability context, agents achieve a functional success rate of up to 90.0%, whereas autonomous reconnaissance, measured by targeted vulnerability recall, plateaus at approximately 50.0%, primarily due to failures in parsing unstructured telemetry. Cross-architectural analysis further reveals distinct capability niches: multi-agent isolation is more effective for long-sequence interactions such as de-serialization, while monolithic and graph-driven designs perform better on short-chain injections and cross-session access-control vulnerabilities, respectively. This decoupled evaluation work provides a fine-grained benchmarking protocol and an empirical basis for designing next-generation automated offensive security agents.

Agentic AI systems powered by large language models (LLMs) are rapidly evolving into autonomous decision-making systems, exposing attack vectors beyond those of traditional LLM vulnerabilities. Existing security evaluations are often tied to specific implementations or domains, limiting unified comparison across heterogeneous systems. To address this gap, we introduce RIFT-Bench, a representation-driven methodology for dynamic red-teaming that enables unified evaluations across diverse agentic architectures. Building on a novel hierarchical representation, RIFT-Bench operates in two automated phases: Discovery, which extracts system structure, and Scanning, which executes adaptive adversarial attacks. It directly evaluates the examined system using 105 adaptive adversarial probes spanning diverse attack vectors and objectives. We demonstrate the effectiveness of the proposed evaluation pipeline across 45 agentic systems spanning a diverse range of implementations, showing that the approach generalizes effectively to heterogeneous agentic architectures. Beyond systems and attacks, RIFT-Bench also supports direct evaluation of mitigation strategies. These key capabilities make RIFT-Bench a scalable foundation for security evaluation of agentic AI systems in practice. Infrastructure code and benchmark artifacts are available at https://tinyurl.com/RIFTBench.

Local LLM agents such as OpenClaw and Nanobot run on end-user machines and act on host resources - the shell, filesystem, browser, stored credentials, and messaging applications - through natural-language goals. These agents have become privileged software runtimes that mediate between user intent, model outputs, and host-level actions. Existing research characterizes the landscape through prompt injection, malicious skills, marketplace risks, or black-box evaluation of agents. But the implementation layer that performs this mediation, the prompt builder, parser, tool dispatcher, skill loader, memory writer, network client, and permission gate, has remained an unexamined safety boundary. To our knowledge, no prior work has examined the agent's source tree to audit these components for implementation-level security weaknesses. We present CLAWAUDIT, a static auditing framework for measuring vulnerability exposure in local LLM agent runtimes. CLAWAUDIT derives a five-category vulnerability taxonomy from STRIDE and develops custom static-analysis rules that target agent-specific patterns absent from established rule sets for vulnerability analysis. We instantiate the taxonomy in two backends, 47 Semgrep YAML rules and 30 CodeQL queries, and evaluate on OPENCLAWBENCH, a benchmark of 446 source-code-level advisories from the OpenClaw repository and split temporally into 229 rule-derivation (train) and 217 held-out (test) advisories. On the held-out test, CLAWAUDIT raises Semgrep recall from 21.7% (Pro baseline) to 66.8%, and CodeQL recall from 13.8% (security-extended) to 75.1%. Train/test gaps remain within 4 percentage points for all four configurations, indicating that the rules generalize to vulnerabilities unseen during rule writing. A preliminary live-code audit shows that these recall-oriented rules require manual triage, motivating semantic filtering before production deployment.

The empirical foundation of cyber deception relies on human-centered hypotheses, but the rapid emergence of autonomous, AI-enabled attackers challenges whether this foundation transfers to AI agents. To address this, we introduce an automated evaluation framework adapted from the Honeyquest instrument to assess LLM attacker judgment at scale. Our 21-LLM cohort spanned 10 providers, diverse architectures and specializations, open- and closed-weight models, and parameter scales from 8B to over 1T. We evaluated the performance of this LLM cohort (yielding 10,962 responses) against the 47-participant human baseline across an identical set of 174 reconnaissance queries. Our empirical evaluation reveals three key findings that establish LLMs as a distinct attacker class: (1) every model in our cohort falls for deceptive traps at a significantly higher rate than human attackers; (2) the defensive attention-diversion effect observed in humans is statistically absent in our LLM cohort; and (3) a critical recognition-action gap, where LLMs successfully articulate trap recognition in their reasoning but exploit the deceptive elements anyway 73.4% of the time. Across the 21 models, trap recognition in reasoning text did not predict fell-for-trap behavior (Spearman $r = +0.08$, $p = 0.73$). Ultimately, these findings demonstrate that human-centered deception hypotheses do not reliably transfer to AI attackers, highlighting the critical need for new research into AI-native active defense frameworks.

As LLM agents increasingly select tools autonomously, their choices among tools with different privileges become safety-relevant. However, prior tool-selection studies focus on safety-agnostic metadata preferences, leaving privilege-sensitive choices underexplored. To address this gap, we study over-privileged tool selection, in which an agent selects or escalates to a higher-privilege tool despite a sufficient lower-privilege alternative. We introduce ToolPrivBench to evaluate whether agents choose higher-privilege tools despite sufficient lower-privilege alternatives, measuring both initial selection and escalation after transient tool failures. Across eight domains and five recurring risk patterns, we find that over-privileged tool selection is common among mainstream LLM agents and is further amplified by transient failures. We further find that general safety alignment does not reliably transfer to least-privilege tool choice, while prompt-level controls provide only limited mitigation under transient failures. We therefore introduce a privilege-aware post-training defense that teaches agents to prefer sufficient lower-privilege tools and escalate only when necessary. Our mitigation experiments show that this defense substantially reduces unnecessary high-privilege tool use while preserving general capabilities.

Memory benchmarks for LLM agents largely assume single-user settings, leaving shared assistants for hospitals, workplaces, campuses, and households understudied. In these deployments, multiple principals write to a common memory pool and query it under different roles, scopes, and relationships, so memory quality requires governance as well as recall. We introduce GateMem, a benchmark for multi-principal shared-memory agents. GateMem jointly evaluates utility for legitimate long-horizon requests with state updates, access control across contextual authorization boundaries, and agent-facing active forgetting after explicit deletion requests. It spans medical, office, education, and household domains, with long-form multi-party episodes, incremental memory injection, hidden checkpoints, structured judging, and leak-target annotations. Across diverse baselines and backbone models, no method simultaneously achieves strong utility, robust access control, and reliable forgetting. Long-context prompting often yields the best governance score at high token cost, while retrieval-based and external-memory methods reduce cost yet still leak unauthorized or deleted information. These results show current memory agents remain far from reliable shared institutional deployment.

Domain-camouflaged injection attacks embed malicious instructions in retrieved content using domain-appropriate vocabulary, evading standard detectors that rely on syntactic injection markers. When detection fails, practitioners need to know which defense architectures reduce attack success. We evaluate five prompting-based defenses (spotlighting, paraphrasing, prompt sandwiching, and two combinations) against domain-camouflaged injection across three model families (Claude Haiku, Llama 3.1 8B, Gemini 2.0 Flash) and three deployment domains (financial, legal, general) using 3,510 trials. Paraphrasing retrieved content before agent processing is the most consistently effective defense in this benchmark, reducing camouflage attack success rate by 55-84% depending on model, and achieves lower attack success rates than our Llama Guard 4 configuration on every model tested. Defense effectiveness is strongly model-dependent: spotlighting halves attack success on Claude Haiku but provides no benefit on Llama 3.1 8B. Financial domain deployments face the highest residual risk at 26-33% baseline attack success rate, with no prompting-based defense fully eliminating the threat on weaker models. These results provide the first systematic evaluation of prompting-based defenses specifically against camouflage-class injection attacks and establish benchmark-based recommendations for practitioners. All tasks use synthetically constructed professional documents; whether these benchmark rankings generalize to real enterprise documents remains an open question.

Tool-using language-model agents introduce security failures that go beyond unsafe text: they can disclose protected objects, write persistent memory, send messages, modify databases, or trigger harmful code and tool effects. Existing evaluations often collapse these stages into a single attack success rate, making it difficult to tell whether a model merely agreed with an attacker or actually produced observable harm. We introduce SafeClawBench, a staged benchmark for tool-using agent security with 600 controlled adversarial tasks across six attack families: direct and indirect prompt injection, tool-return injection, memory poisoning, memory extraction, and ambiguity-driven unsafe inference. SafeClawBench reports three separate endpoints: semantic attack acceptance, audit-visible harm evidence, and sandbox-observed tool/state harm. Evaluating five agent endpoints under four prompt-level policies, we find that these endpoints capture different failure modes. Without additional prompt protection, semantic failure rates vary widely across models, from 9.0% to 44.2%. Audited harm evidence is narrower than semantic failure, and under a separate executable protocol some matched task identities produce sandbox harm despite passing the Semantic Core call: in a 12,000-row matched analysis, 291 of 347 observed sandbox harms occur in rows that pass the semantic check. Prompt policies change endpoint outcomes, but their effects depend on both model and protocol. SafeClawBench provides a reproducible framework for comparing agent models and prompt-policy conditions without conflating textual compliance, evidence-supported harm, and executable state changes. The open-source dataset is available at https://huggingface.co/datasets/sairights/safeclawbench.

Prompt injection defenses evaluated on synthetic benchmarks do not generalize to real enterprise documents, which are longer, denser, and interleave legitimate authority language with factual content. We demonstrate this gap with a real-document benchmark of 122 tasks across five professional domains (financial, legal, medical, scientific, DevOps) using actual SEC filings, Federal Register rules, PubMed abstracts, arXiv papers, and GitHub postmortems. Paraphrasing, the strongest defense on synthetic benchmarks, shows no statistically significant attack success rate reduction on real documents (p=0.500) while degrading utility from 91.8% to 82.8%. We introduce PARSE (Provenance-Aware Retrieval Sanitization), a domain-aware, fact-preserving sanitization pipeline that classifies each sentence by injection likelihood, extracts structured facts before rewriting, and verifies fact preservation via a consistency-checking loop. A directiveness gate routes 59% of real enterprise documents to a lightweight path, concentrating computational cost on high-risk documents. PARSE achieves 15.6% attack success rate -- a 38% reduction versus the 25.4% baseline -- at 86.9% utility, the only condition that is both statistically significant (p=0.014, adequately powered) and maintains near-baseline utility. Practitioners should evaluate defenses on domain-matched real documents, not synthetic proxies.

Open-source LLM agent ecosystems are growing rapidly, yet the security of community-contributed skills - modular tool definitions that extend agent capabilities - remains largely unvetted. The gap we fill: existing scanners operate at the code layer and are structurally blind to instruction-layer and multi-agent risk - natural-language directives that hijack an agent, exfiltrate data through encoded side channels, or chain harm across pipelines - so what is needed is a semantic, multi-dimensional vetting system rather than another signature matcher. We present SKILLVETBENCH, a live public leaderboard on Hugging Face that uses an LLM-as-Judge to vet agent skills. What is new: SARS (Skill Agentic Risk Score), a five-dimensional agentic-risk metric with a principled weighted formula for instruction-following systems. What is integrated: full CVSS v4.0 vector decomposition and a ClawHub dual-view that places our LLM-generated review beside the official marketplace verdict. What is demonstrated: drawing on our companion benchmark paper [ 1], the LLM-as-Judge stage achieves zero false negatives across 78 confirmed-malicious skills and zero false positives across 22 benign controls, while the best static baseline (SKILLSIEVE) still misses 15%; for instruction-layer categories such as Prompt Injection and Memory Poisoning, conventional tools miss between 89% and 100% of threats (e.g., CODEBERT detects none of nine memory-poisoning skills). Detection rates vary from 35% to 95% across four LLM evaluators, motivating ensemble scoring in production deployments.

We ran 300 repeated vulnerability-finding scans to measure how repeatable agentic large language model (LLM) security review is on the same JavaScript code, prompt, and benchmark harness. The headline result is that LLM security findings were unevenly repeatable: reference-matched findings were stable, but extra model reports varied heavily from run to run. Across 250 model runs, 80 of 161 unique unmatched findings appeared in only one of five identical repetitions, while only 22 appeared in all five. By contrast, when Claude matched a Snyk Code reference finding, the behavior was much more stable: 134 of 158 unique reference-matched findings appeared in all five repetitions. The benchmark also shows complementarity. Models consistently found familiar, high-signal exploit shapes, and in one case surfaced a likely Snyk Code product gap. Snyk Code static application security testing (SAST) was deterministic and better at systematically enumerating repeated data-flow sinks. The results support combining agentic LLM review with deterministic SAST rather than treating either technique as a replacement for the other.

Skills are becoming the capability layer through which LLM agents turn plans into actions, but their use introduces security risks such as data leakage, unauthorized operations, and tool misuse. Existing vetting usually evaluates each skill in isolation, while real agent tasks often invoke multiple skills in a shared execution context. This creates Skill Composition Risk (SCR): a skill that appears benign alone can become harmful when its outputs, trust signals, authorization cues, or side effects influence later invocations along an activated path. We introduce SCR-Bench to evaluate this risk in controlled, sandboxed skill environments. Rather than relying only on textual intent or surface behavior, SCR-Bench records downstream state changes and path-level outcomes across composed skill executions. It contains three sub-benchmarks: SCR-CapFlow for capability-flow composition, SCR-TrustLift for trust-transfer composition, and SCR-AuthBlur for authorization-confusion composition. Across SCR-Bench, composed paths expose risks that are largely absent under isolated evaluation. In SCR-CapFlow, attack success rate reaches 33.6 percent under composition, compared with near-zero isolated baselines. In SCR-TrustLift, attack success rate exceeds 96.5 percent on four of five backends. In SCR-AuthBlur, the risky-approval rate increases by 71.8 percent relative to the L0 isolated baseline under the L1 context setting. These results show that agent skill security should be assessed at the level of activated paths rather than isolated artifacts. SCR and SCR-Bench provide a foundation for path-aware risk evaluation and defense in LLM agent skill ecosystems. Benchmark: https://github.com/saint-viperx/SCR_Bench.