Tool-Using Agents

Coding harnesses shape how autonomous coding agents translate model capabilities into long-horizon software-engineering performance, yet existing work typically evaluates harnesses as monolithic systems, leaving the effectiveness of individual components unclear. To enable component-level comparisons, we study this question with a lightweight coding harness whose execution loop is fixed while three components are varied: planning, action space, and context management. Across four models evaluated on SWE-Bench Verified and Terminal-Bench 2.1, we evaluate 176 matched settings spanning five context-management strategies, four context-window budgets, and targeted ablations of planning and action space. We find that: (1) Context management becomes increasingly valuable as the context-window budget tightens, with most of its benefit coming from preventing context-overflow failures. (2) Staging rule-based elision before LLM-based summarization provides the strongest overall efficiency among the context-management strategies, whereas making elided content recoverable adds machinery that models rarely use and yields no accuracy gain. (3) Planning shifts from an accuracy scaffold for weaker models to a cost saver for stronger models, with little change in accuracy. (4) Predefined tools improve performance for models with weaker bash proficiency, whereas bash-capable models can operate effectively with a bash-only interface and achieve substantially lower cost, especially on command-line-centric tasks. Trajectory-level analysis explains these effects: context management extends execution trajectories without substantially altering agent behavior, planning changes where trajectories stop, and the action space changes the granularity at which code is written. These findings inform model- and budget-aware harness design and provide a modular framework for evaluating future harness components.

An agentic request spends substantial wall-clock time waiting for tools, and its KV cache holds GPU memory the whole time. Serving systems decide whether that cache stays, leaves, or comes back by guessing how long the tool will run, from the tool's name, its history, a duration declared before the call, or the engine's own occupancy. We show that no estimate fixed before a call starts can know its duration, and such estimates may not even rank the calls. Meanwhile, the running tool already holds the answer, but the agent stack together with the tool silences it. We propose that tool calls report their progress explicitly while they run, and we measure what that takes. A census of four public agent corpora finds a readable signal in most tool time once it is revealed, in two strengths: a fraction of the work remaining, or an accurate signal that the end is near. A harness recovers it without changing what the agent sees, at no measurable cost to the agent's benchmark score. At the points where a KV cache decision is made, the reported progress is between several times and an order of magnitude more accurate than the best published predictors, and it stays accurate when the environment changes. Plugged into a production engine through a few small hints, it cuts the p90 time to first token (TTFT) after a tool call by 20.7% (HBM only) and 20.8% (HBM + DRAM) against LRU, close to an oracle. A serving system should not guess what its tools can tell it.

Machine-learning libraries such as Weka and scikit-learn were designed for human programmers. Language-model agents now use these same libraries by recalling APIs from memory and writing code, an approach that hides what a library offers, delays errors until runtime, and loses experimental state between turns. We present TuiML, a self-contained machine-learning library built for AI agents, with native algorithms across supervised, unsupervised, time-series, data handling, tuning, and evaluation tasks. Every component describes itself through machine-readable metadata and parameter schemas, so an agent can search the library, inspect components, compose validated workflows, and register new ones that become discoverable in turn. Every call is validated, seeded, and traced, and sessions export as runnable notebooks, making experiments reproducible by construction. One specification layer drives the Model Context Protocol (MCP), agent-framework adapters, a Python API, a CLI, and local model serving, while data and models never leave the machine. Benchmarks show TuiML remains predictively competitive with scikit-learn and Weka. While looking like a conventional library to a human user, TuiML is designed for agents first, allowing them to read, extend, and operate machine learning autonomously. TuiML is open source, with documentation at https://tuiml.ai.

AI agents increasingly execute long-running workflows that externalize effects through independently supplied tools. Under retries, speculative execution, concurrency, and partial failures, the resulting external state may be inconsistent with the workflow's intended resolution: required effects may be missing or duplicated, aborted effects may survive, and committed effects may depend on provisional state that is later withdrawn. Advanced transaction models address related failures, but assume that lower-level operations expose the semantics they depend on: whether an effect occurred, whether it can be compensated, staged, or safely reordered. Shared agent-tool interfaces usually do not. We contribute an effect-history model that separates events in the external world from the runtime's observations of them, and a catalog of eight recurring external-effect anomalies. From the catalog we derive the boundary capabilities required to exclude each anomaly in general, and four points where black-box tool invocation alone cannot provide a general guarantee. We then ask how much of this is expressible in a widely used shared tool interface, measuring the use of the standard annotation vocabulary across 98,291 tools exposed by registered Model Context Protocol (MCP) servers. The fields are widely emitted but provide only coarse call-level hints, and none of the required capabilities is fully expressible. These results motivate reusable transactional contracts at the tool boundary.

Memory-augmented and tool-using agents expose exact private values when remote LLMs process retrieved memory, tool actions, and intermediate observations. One-way masking limits direct exposure but removes values needed for trusted execution and can leak them through later observations. We propose Trustworthy Virtual Memory (TVM), a closed-loop runtime that keeps exact-value state local while presenting a protected view to the remote model. Within this single runtime, Rule-TVM replaces whole protected fields with locally recoverable handles, and Semantic-TVM instead replaces only sensitive spans predicted by a trusted local model, preserving surrounding task-relevant context. On Memory-EHR and Memory-RAP across two providers, span-level projection recovers most of the EHR utility lost under whole-field replacement (Task Success 84.17% vs. 52.33% on DeepSeek) while measured exposure stays low and workflows remain executable.

Existing agent benchmarks mainly evaluate final task success or tool-call correctness, providing limited insight into whether agents can reliably diagnose and recover from intermediate execution failures. This limitation becomes particularly critical in multi-turn parallel tool-use scenarios, where errors may propagate across dependent branches and trigger cascading failures. We introduce ParaRecover, a process-level benchmark for evaluating error localization and recovery in multi-turn parallel tool-use agents. Built upon a fine-grained taxonomy of 14 error types covering planning dependencies, tool selection, and argument matching, the benchmark comprises 10,626 instances spanning two difficulty levels. To enable finegrained, process-oriented evaluation, we further propose the SDE rubric, which measures structural integrity, diagnostic reasoning, and evolutionary strategy during agent execution.Experiments across more than ten mainstream LLMs reveal that even state-of-the-art models still struggle with multi-turn error propagation,implicit tool-use failures, and precise replanning. Moreover, we demonstrate that the SDE rubric provides effective supervision signals for improving agents' reflective recovery capabilities. Our data and code are available at https://github.com/gbw206/ParaRecover.

Real-world Earth observation (EO) agents must translate high-level scientific questions into executable workflows to acquire observations, prepare data, perform domain computations, and derive conclusions from runtime evidence. Existing EO agents typically start from supplied observations, while benchmarks typically provide prepared inputs or candidate answers, leaving full-chain open-world EO execution largely untested. We present Earth-Agent-Pro, an execution-adaptive Plan-and-Execute framework using expert-authored skills to constrain planning and runtime tool use. Workflow-centered structured memory records planned steps, accepted evidence, and their dependencies, enabling repair of only the affected workflow suffix when runtime evidence invalidates a step. Separate large language model adapters use sequence-level supervised fine-tuning for planner workflow composition and node-level group relative policy optimization with locally verifiable rewards for executor tool-argument grounding. Earth-Bench-Pro instantiates 248 expert-curated task cores as 744 questions under three matched regimes. Its 248 Open-World Execution questions span RGB imagery, spectral observations, and remote sensing products, pairing high-level requests with runtime data requirements, executable trajectories, and open-ended answers grounded in execution evidence. With a shared GPT-5 backbone, Earth-Agent-Pro achieves 66.13% LLM-as-Judge accuracy, exceeding ReAct by 20.95 points in this metric and 24.44 points in Tools-In-Order. Joint adapter tuning raises Qwen3.5-9B LLM-as-Judge accuracy from 38.31% to 50.00%, an 11.69-point gain over the untuned configuration. Planning-only evaluation and execution with the reference workflow show that the adapters improve workflow composition and argument grounding, respectively. Code and datasets will be released soon.

Tool-augmented language models are evaluated on whether they reach the right answer, not on whether they report honestly when a tool fails to supply one. We isolate this post-failure decision with a benchmark of 1,024 items spanning 16 internal-system domains and eight tool-failure types, in which a tool call is enforced and the returned payload is guaranteed to be unusable. Under a deployment-style system prompt, 14.10% of responses are dishonest: the model either asserts a value the payload cannot support or declines while citing a fabricated policy or capability limit. The rate is governed almost entirely by whether the failure is signalled. When the tool returns status:error, dishonesty is absent (0.0%); when it returns status:ok with a redacted, corrupted, stale, malformed, empty or truncated value, dishonesty reaches 45.3%. The behaviour is not an artefact of our prompts: it appears under a neutral prompt (10.17%) and under the shipped prompt of every production agent framework we evaluate, reaching 24.67% under CrewAI's, and none of the nine frameworks we audit specifies what the model should do when a tool fails. Comparing prompt-level defences, we find that the operative variable is not deference to tool output but the absence of a named failure state. Appending a single sentence that requires the model to emit retrieval_status: OK or FAILED before answering reduces dishonesty from 14.10% to 0.87%, with one item of 688 worsening against 92 improving, and transfers unchanged into three foreign agent scaffolds. The emitted flag is faithful in 99.7-99.9% of declarations, giving a runtime detector that needs only a regular expression.

Language models act through tools, yet practical agents face libraries containing thousands of interfaces. We introduce the tool menu as the short, ordered subset of available tools shown to an agent before execution. The agent can call only tools in this menu. Multi-step tasks require the final action and the prerequisite tools that create its inputs in a usable order. Current constructors rank tools by request relevance, which can surface the final action while omitting or delaying less obvious producers. We introduce the state path, a pre-execution route from the observable request state to the desired outcome, and propose State-Path Tool Menu to learn it. Our framework treats the menu as an execution prior over these routes. Its encoder represents which tools can run from the current state, how their outputs satisfy later inputs, and which orders recur in training paths. A retriever covers an executable entry, the missing-input producers, and the final action. A reranker then places producers before consumers. On ToolBench, our menu raises online success from 0.737 to 0.898 and outperforms retrieval, reranking, generation, and routing baselines without changing the agent. The State-Path menu also covers more complete chains with 32 tools than the official list covers with 128, and its success gain persists across executor families with different model capacities. Our code is at https://github.com/Met2348/State-Path.

Long-running AI agents may read state, reason, wait for tools or human approval, and perform an external action much later. The state that justified the action can change in the meantime. For example, after an agent proposes an 80 GBP refund under a limit of 100, a customer-name change affects only presentation metadata, a new limit of 90 still permits the refund, a limit of 50 invalidates it, and a refund issued by another worker must prevent a duplicate. Standard optimistic concurrency control and version checks can detect that previously read state has changed, but by themselves do not determine whether that change invalidates the pending action's justification. We call any detected version change a version conflict; when that change invalidates the action's justification, it is also a decision conflict. ATR records the explicit, executable conditions that justify a pending action and rechecks only the conditions affected by a change before releasing the external operation. It can retain the action, refresh non-decisive metadata, require replanning, or block execution; a target-side transaction or compare-and-set binds checked state to commit. Across 210,000 controlled executions over 15 mutation cases, ATR matched every developer-specified outcome with no false allows or blocks. In ten durable SQLite checkpoint/resume cells, it evaluated 0.6 conditions per change versus 6.0 for FullScan. At 4,093 recorded reads, ATR took 9.3 microseconds versus 2595.9 microseconds for FullScan. These deterministic results establish controlled feasibility, not production generality or automatic extraction of the required conditions.

Conversational assistants can blend retrieval, action selection, escalation, and wording in a single model path, or separate those roles. We report a production migration of a customer-support assistant at a large accommodation marketplace (millions of conversations per month, 11 languages, 10-second P90). Dynamic Response (DR) replaces a single Qwen3-235B-A22B blended responder with a bounded ReAct orchestrator over typed tools plus a smaller generator that writes from a backend-validated context contract. Because the migration also changed prompts, alignment, and serving, we attribute each effect to its cause and claim as architecture effects only those measured on identical replayed turns: typed entity selection moves the reservation selector to a precision-first operating point (precision 8.3% to 89.1%, recall 75.2% to 67.3%), and typed action IDs with a membership check remove observed structured-action hallucination (2.14% to 0.0%). A low-ramp A/B test reproduces the replay escalation reductions: hard-escalation responses fall from 5.60% to 3.08% and soft-escalation responses from 9.56% to 2.49%, while production handoff volume holds roughly steady; self-solve is directional (+5.1 points, 95% CI [-2, +12]). Serving optimizations cut orchestrator P90 latency from 3.87s to 2.24s on a GPU footprint reduced by roughly one-third, and self-hosting reduces estimated annual model-serving cost by more than an order of magnitude.

Tool-using LLM agents spend wall-clock time not only on model inference but also in serial action--observation turns, where each tool call, environment transition, and observation can delay subsequent decisions. We introduce \textbf{Speculative Macro Commit} (SMC), a runtime mechanism for a two-tier agent system: a large authoritative actor model produces the official trajectory, while a faster speculative drafter model continuously predicts and executes future action chains on an isolated environment snapshot. SMC mines recurring multi-action skeletons from training traces and stores them in a macro library used to match against action chains predicted by the drafter at runtime. When the actor's next tool call matches the first drafted action, SMC commits the remaining pre-executed draft steps, together with their observations, to the official trajectory. Using Qwen3.5-27B INT4 as the authoritative actor model and Qwen3.5-4B as the speculative drafter model, SMC matches the sequential agent's overall accuracy while reducing latency by 10.23% over the Speculative Actions (SA) baseline and 18.59% over sequential execution on the $τ^2$-Bench Telecom subset. On AppWorld, SMC reduces wall time by 7.7% over SA baseline and 44.9% over sequential execution, with a small reduction in task completion. Overall, SMC provides a practical way to reuse multi-step speculative execution and reduce agent latency beyond single-step speculative actions. Our code is publicly available \href{https://github.com/zeyuliu1037/speculative-macro-commit}{\textcolor{magenta}{here}}.

Tool-using agents are being proposed for medical imaging, and their behaviour when a tool returns a false finding is largely unmeasured. We audit whether a ReAct-style tool-calling agent abandons an answer it has already given correctly once a falsified finding arrives, and whether that depends on how the finding is presented. On 20 VQA-RAD closed questions across four vendor-designated model tiers, the agent commits to an answer from the image alone; a negated finding is then delivered either as JSON from an analyze_image tool the agent invokes itself, or as quoted prose attributed to a radiologist. Our outcome is the commission-error rate over cases answered correctly without any tool. Deference is much higher under the prose-attributed claim: at the strongest tier the agent revised its correct answer in 10 of 13 cases against 1 of 13 under the tool (exact McNemar p=0.0039, Holm-adjusted 0.012). We do not claim this isolates the source label. Attribution travels with the delivery channel in our design, and exposure differs because the tool claim reaches the agent only when it calls the tool. The finding is a joint source-and-delivery asymmetry from a small-scale pilot whose pre-specified stopping rule was not met.

Tool-using agents increasingly rely on external tools to complete multi-step tasks, but tool returns can fail in different ways and require different recovery actions. Existing robustness studies often use uncertainty-based measures to detect when an agent becomes unreliable. These measures can reveal that something has gone wrong, but they do not directly identify the type of tool failure or the appropriate response. We address this limitation by analyzing tool failures at the moment a return enters the agent context. Our approach combines two complementary signals. The first compares the likelihood of the returned content under the tool schema and under the full trajectory prefix. The second measures the agent's probability distribution over its legal next actions. We evaluate the approach by injecting incomplete and inconsistent returns into a retail customer-service benchmark. The results show that likelihood-based signals clearly capture incomplete returns and some direct inconsistencies, while action-based signals reveal how strongly a failure changes the next decision. Some failures that are weak under likelihood signals can still redirect the agent toward state-changing actions. These findings show that tool failures can be recognized at the return boundary, but reliable diagnosis requires combining multiple signals.

Recent agent benchmarks increasingly ground evaluation in executable environments, from code repair to web navigation, app APIs, and function calling. Yet completing consequential work beyond code requires more than producing a plausible response or valid tool call: agents must gather missing information over multiple turns, follow domain policies, coordinate dependent tools, and realize the correct persistent state transition without collateral effects. In this paper, we introduce Thinkingbox, a sandbox for tool-agent-user interaction that provides isolated MCP-compatible tool sessions, complete execution traces, and outcome evaluation over terminal backend state. Built on this sandbox, Thinkingbox-bench contains 507 policy-conditioned workflows across business scenarios, including retail, hospitality, auto insurance, neobank internal IT, and consulting IT/HR support. Each attempt is evaluated by task-specific executable checks that accept valid trajectories while rejecting wrong, missing, or extra effects; designated tasks additionally check required properties of the final response. Our experiments reveal that even the strongest proprietary and open-weight models show steep reliability drops: Claude Opus 5 falls from 66.50% pass@1 to 47.53% pass^20, and Kimi-K3 from 57.37% pass@1 to 17.60% pass^20. Moreover, many failed trials terminate cleanly after valid state-changing actions, so response- or tool-call-level signals poorly proxy end-to-end completion. Thinkingbox-bench reveals a large gap between occasionally finding a successful trajectory and reliably completing stateful business tasks. We release both Thinkingbox (https://github.com/microsoft/thinkingbox) and Thinkingbox-bench (https://github.com/microsoft/thinkingbox-data).

Tool-using agents consume external data from sources with different levels of trust, yet tool responses rarely identify who produced each component or what it should convey. We show that this gap enables state-corruption attacks, in which attacker-controlled content makes environmental claims beyond the informational authority of its response component and corrupts the agent's perceived environment, making the resulting action appear justified to existing guardrails. We introduce PIPES (Provenance-Informed, Prior-Enforced Screening), which screens response units using semantic priors and source provenance. PIPES uses static field contracts when schemas provide stable expectations, and conditions screening of open-ended content on the pre-response trajectory and trusted provenance metadata. It marks units that violate their semantic prior or the provenance hierarchy; deployments may remove, warn, block, or escalate detected violations. We instantiate atomic removal and evaluate PIPES against adaptive PAIR-style attacks. Across the three VitaBench and three AgentDyn splits with Gemma 4 31B IT as the target agent, PIPES reduces average attack success from 84.7% to 2.3%, while preserving average benign utility (92.5% with PIPES versus 90.6% without defense).

Long-running LLM agents act through tools, and a single step can send an email, merge a pull request, or wire a payment. The steering decision is the pre-commit choice at that boundary: proceed, or hold for human or policy review. We introduce SteerBench-Work, an incident-anchored, bidirectional benchmark for that decision in workplace agents across developer operations, customer service, finance, legal, medical, HR, and security. Release v2026-05 contains 106 scenarios anchored in public incidents, paired evidence-reversed mirrors, and calibration controls, with labels split nearly evenly between proceed and hold so the two error directions get near-identical numbers of chances. A model sees the proposed action and the available evidence, returns a gate decision, and is scored on whether it crosses or holds the boundary correctly. Across 30 model conditions the failures run almost entirely in one direction: models wrongly hold authorized, evidence-cleared work on 28.1% of opportunities and wrongly allow unsafe work on 1.0%. The hardest cases are risk-resolved commits, where signed or structured evidence has already cleared a real risk trigger, and models score markedly worse on evidence-reversed mirrors of famous incidents (63.8%) than on the incidents themselves (98.5%). General capability is not the same as steering calibration: higher-capability models often over-refuse at the commit boundary, and more reasoning can repair a weak gate while leaving a calibrated one flat. The public leaderboard is at steerbench.com.

When a tool-using agent is given the same task in a different language, does it still take the same steps? Multilingual evaluation rarely asks: it compares final answers and discards the actions. Yet those actions are the product: they fix cost and latency, decide how the system fails, and are the only auditable part of its behaviour. We make the action policy the measured object across 8 models, 6 parallel benchmarks and 41 languages (2.38M rollouts). The naive measurement fails: five confounds sit between raw trace similarity and any defensible claim, each able to flip a conclusion. Short traces score higher, empty traces score perfectly, unrelated traces agree by chance over half the time, the gap is capped by each model's reproducibility, and a model asked the same question twice in one language answers differently, leaving no baseline. We remove all five, and every correction makes the effect larger. Divergence proves structural, not sampling noise: it survives greedy decoding in every cell and stays flat as temperature rises, even as models grow less self-consistent. Normalised by their own reproducibility, four very different frontier models converge under greedy decoding, each keeping 71-73% of its action policy across languages, with model identity explaining only 5.7% of the variance. Below roughly 10B parameters it breaks down, and the ordering among smaller models is largely an artifact of a chance floor we measure by permutation rather than assume. Agents route non-English tasks through English; this pivot is causally load-bearing, confirmed by a pre-registered prediction across four models, and models will not abandon it when told to. Finally, a single trace-extraction regex, not the model, manufactured a multilingual failure: two worked examples raise one model's measured accuracy twenty-sixfold while its accuracy on readable outputs barely moves.

Agentic workflows have become an important abstraction for building reliable LLM-based automation systems by organizing large language models (LLMs), tools, and control logic into explicit execution structures. However, constructing high-quality agentic workflows remains largely manual and requires substantial domain expertise. Recent studies have explored automatic agentic workflow generation from historical task-solving records, but they mainly produce LLM-centric workflows, where real tool executions are abstracted and simulated by LLM nodes, limiting the usability and stability of generated workflows. To address these limitations, we propose FlowScout, an execution-guided framework for generating tool-integrated agentic workflows from historical task-solving records. Specifically, FlowScout represents an agentic workflow as a directed graph composed of LLM nodes, tool-calling nodes, and dependency edges. It first mines a common tool coordination skeleton from historical records to construct an initial workflow, and then refines the workflow topology through Monte Carlo tree search guided by execution feedback. We evaluate FlowScout on four representative task domains and compare it with three baselines, i.e., PM4Py, ReAct and AFlow. Experimental results show that agentic workflows generated by FlowScout improve tool invocation correctness by at least 92.69% and execution quality by at least 17.66% over the baselines, while achieving lower performance variation across repeated runs.

Large language models (LLMs) increasingly rely on external tools to accomplish complex real-world tasks, making tool documentation a critical grounding resource for LLM agents. Existing studies mainly focus on improving the tool-use capabilities of LLM agents, while largely treating tool documentation as a fixed input. Although several recent works attempt to optimize tool documentation through rewriting or compression, little is known about how the information contained in tool documentation affects agent performance across different settings. To bridge this gap, we conduct a large-scale empirical study on tool documentation for LLM agents. Our study reveals substantial heterogeneity in the information fields provided by existing tool documentation. Moreover, the effectiveness of different information fields is highly dependent on the task domain, LLM backbone, and agent paradigm, indicating that no fixed tool documentation can consistently generalize across diverse agent settings. Motivated by these findings, we propose DocsChisel, an adaptive tool documentation optimization framework for LLM agents. DocsChisel analyzes failed execution traces of a target LLM agent to identify documentation-related issues, and iteratively optimizes tool documentation by adding, removing, and refining information fields for each tool. We evaluate DocsChisel against two state-of-the-art baselines, i.e., EasyTool and DRAFT. Experimental results show that DocsChisel improves the task success rate of LLM agents by 95.89% over the original tool documentation and by 75.15%, on average, over existing baselines, while incurring limited optimization time and token overhead

Agentic benchmarks aim to measure how well AI agents plan, search, execute, and recover within realistic multi-tool environments, but they are almost exclusively in English. As AI agents are globally deployed to a linguistically diverse user base, whether agentic competence measured in English transfers to other languages remains an open question. We introduce OmnilingualGAIA2, a machine-translated expansion (with partial human- expert validation) of the GAIA2 agentic benchmark, covering ten target languages spanning five writing systems, paired with a localised and human-calibrated multilingual verifier. Evaluating seven frontier and open-weight agents, we find a universal cross-lingual gap of 8.8-18.4 pass@3 points that is agent-asymmetric in magnitude, concentrates on tool-orchestration rather than quantitative reasoning, and does not close with model scale. A stratified error attribution decomposes the gap as predominantly model-driven (55%), with a bounded translation-contamination floor of only 6.4% of scenario-language pairs. Human-expert linguistic analysis further identifies morphological cue loss and amplified ambiguity as the primary failure mechanisms in non-Latin-script languages. Our results argue that multilingual agentic evaluation must become a standard part of the reporting protocol for globally deployed agents.

How much an AI coding agent costs to run can depend more on the agent scaffolding that drives it than on the interface through which it reaches its tools. We set out to measure the cost of tool use over the Model Context Protocol (MCP) against tool use over an ordinary command-line interface (CLI), a difference on which published estimates disagree by more than an order of magnitude while resting on practitioner reports that cannot be reproduced. We ran one fixed software task -- six operations against a private online git repository -- across seven agent scaffoldings and five language models, and we verified completion by inspecting the repository state rather than trusting the agent's self-report. The dominant effect was the scaffolding. Two of the seven ship no MCP support at all; they completed every run using only the CLI, which shows that MCP is unnecessary for this class of work, and they were 5.0x to 28x cheaper than the five scaffoldings that do support MCP, comparing CLI runs alone with no MCP server attached anywhere. The effect was largest for a small 27-billion-parameter model running locally, whose cost varied 139x across scaffoldings while it completed the task under all of them. The comparison we set out to make proved unstable: thirteen strictly paired MCP-to-CLI ratios span 0.43x to 29x, with outliers on both sides. The two interfaces separate on the cost of failure, where 12.9 per cent of the money spent on MCP runs bought no completed work against 2.2 per cent on CLI runs, but not on its frequency: failures were equally common in both, in the original runs and in their repetitions alike. Agents frequently ignored the interface they were assigned, so comparisons that do not verify actual behaviour measure an unknown mixture. The harness, the task, the verification and the complete dataset are released as open source.

Delegating a web task involves more than asking a question; it requires transferring a policy: what to verify, how to handle uncertainty, which preferences matter, and when to stop. Yet, current live-web agents are evaluated solely on the final answer, ignoring the policy constraints that define the delegation. A plausible final answer can conceal violations of that policy. Our full live audit reveals this critical gap: a strong controller completes 99.2% of tasks but honors all policy constraints in only 38.8% of cases. Finishing does not imply fidelity. WebRider bridges this gap by formalizing the delegated policy as an intent contract---an operational record of goals, constraints, evidence obligations, answer form, and task-local persona controls that must hold even as web pages change. WebRider employs a hierarchical architecture: a top-layer controller maintains the contract, a middle layer realizes intentions as guarded executable actions, and a tool layer executes these actions via browser, search, and maps tools. Our benchmark, RiderBench, evaluates this design on 4,096 live-web contracts across 42 public websites, auditing both the internal contract state and the visible user experience to determine if a rollout preserved its policy and if the steps were persona-consistent. The guarded middle interface also serves as a high-quality training signal; an 8B action-policy model trained through this interface outperforms executable-only baselines under a fixed controller. By making the browsing path a first-class object, WebRider enables a system that is auditable, human-judgeable, and learnable without conflating action realization with final-answer decisions. Dataset URL: hf.co/datasets/WebRider/WebRider.

Prompt-side playbooks can improve tool-using language agents without retraining, but their portability beyond the source setting is unclear. We study frozen playbook transfer under a shared distill--validate--transfer protocol. On ALFWorld, transfer is beneficial under controlled greedy decoding and, in one near-budget-matched comparison, distilled guidance outperforms five fixed demonstrations. On TAU2-Bench, a prespecified aggregate contrast supports a modest average matched-domain advantage, but global Holm correction retains only one of 135 route-level effects; the remaining grid provides descriptive evidence of compatibility-sensitive heterogeneity. On XBench-DeepSearch, one artifact--runtime pairing preserves useful first-try heuristics while producing repeated queries, delayed stopping, and substantial cost inflation after a context-runtime shift. Across benchmarks, transferred and target-derived playbooks both require target-side validation of success, termination, protocol compatibility, and cost. Frozen transfer is therefore a conditional cold-start option, not a reuse-by-default strategy or a universally preferable alternative to target-side redistillation.

When will robots have their ChatGPT moment? Such a breakthrough requires a general-purpose robot that can handle unfamiliar tasks in unfamiliar environments, remain controllable over long interactions, and learn from experience. Today's embodied systems largely follow an end-to-end observation-to-action path. Despite rapid progress, they remain far from this goal: their generalization depends heavily on the coverage of robot training data, while long task execution remains difficult to control and inspect. To realize this goal, we introduce the Embodied Task Agent (ETA), a new paradigm for extending digital agents into the physical world, and release OpenETA as its open-source implementation. ETA centers the robot around a Planner that chooses one Tool call at a time, an Interface that controls execution, and a World that returns the result and a fresh observation. This loop allows the agent to verify outcomes, adapt its plan, and turn successful and failed interactions into reusable experience. OpenETA provides replaceable Planners, composable Tools and Skills, auditable memory, replayable trajectories, and common interfaces for simulation and real robots. For Codex, OpenETA can operate as a lightweight plugin that exposes only observe, mark_point, and move_to.

Onboard satellite intelligence requires a task layer that translates mission intent into local tool calls, exposes execution state, and returns machine-consumable artifacts under communication and power constraints. We present SAT-Edge-Agent, a hardware-in-the-loop (HIL) edge-agent system deployed on a commercial off-the-shelf ARM-based heterogeneous edge system-on-chip. A browser workspace and FastAPI agent coordinate a local OpenAI-compatible language service with a project-internal YOLO-style oriented-object-detection endpoint that returns FAIR1M metadata-backed structured results. Two fixed FAIR1M workloads, one single-image and one serial two-image request, were repeated 20 times each and completed 20/20 attempts. Mean Full-Agent latency was 29.353 s and 60.937 s, with empirical P95 values of 31.166 s and 66.882 s. Mean detector time was 861.386 ms and 1510.920 ms, only 2.93% and 2.48% of the corresponding Full-Agent means. Profiling indicates that most visible latency occurs outside detector execution. Mean CPU utilization was 20.761% and 20.482%. A 200-ms NPU-load field averaged 100% for both workloads, but it represents a shared-accelerator software field rather than detector-only occupancy or calibrated utilization. The public evidence package provides sanitized request-level records, redacted JSON, normalized SSE examples, and scripts reproducing the reported statistics. These results establish a reproducible HIL boundary for observable satellite edge-agent orchestration, but do not establish detector accuracy, a new geolocation method, calibrated energy efficiency, or flight readiness.

Recent advances in persistent personal-agent frameworks are making human-centered agent networks realistic deployment targets: each user can be served by an AI agent that acts on the user's behalf, maintains state, and communicates with other agents through social and task relations. In these networks, everyday tool use becomes multi-party owned-agent collaboration over personal workspaces, where files, records, tools, and policies are not directly visible across owners. Existing agent benchmarks study tool use and collaboration, but they do not provide an end-to-end sandbox for verifiable cross-user agent collaboration with realistic user digital workspaces or test how harmful actions can travel through the human-centered agent network. We introduce WeClawArena, an auditable benchmark and runtime sandbox for multi-party owned-agent collaboration over personal workspaces. WeClawArena targets collaborative tool-use tasks in which personal workspaces serve as both operational tools and personal constraints. The benchmark contains 124 base tasks across six cross-user task domains and expands them into 620 scenario variants, with one benign control and four attack-vector variants per base task. The sandbox records peer messages, tool calls, resource operations, governed decisions, and final workspace states. WeClawArena reports utility and attack success rate separately and audits attack success from bounded runtime evidence, supporting diagnosis of task breakdown, privacy leakage, poisoned evidence, and invalid authority paths.

The performance bottleneck of agents is increasingly shifting from model capability to the robustness of their execution processes. Tools play a central role as the primary interface through which agents interact with external environments, yet existing methods rarely focus on ensuring robust tool use across diverse runtime conditions. To address this problem, we propose ExpG, a mechanism that builds and refines adaptive guidance capturing each tool's capability boundaries and best practices, thereby enabling agents to use tools more robustly and effectively. ExpG consists of three phases: (1) experience acquisition, which analyzes tool invocation quality from historical execution trajectories, producing structured learnable experiences through multi-aspect attribution; (2) experience distillation, which keeps the experience pool effective by filtering unhelpful experiences, selecting representative ones with an equivalence-class-based method, and summarizing them into generalizable guidance; and (3) experience reuse, which applies the guidance adaptively during future task solving. Extensive experiments show that ExpG brings consistent improvements across the tool selection, tool calling, and response generation tasks, enabling smaller agents to outperform larger ones that do not use ExpG. Moreover, ExpG achieves particularly strong gains in challenging settings, suggesting a promising path toward more robust tool use. Our code, experiments, and results are available.

To operate robustly in open-world environments, autonomous agents should be able to infer the behavior of unfamiliar systems through interaction alone, even in the absence of documentation. However, existing tool-use benchmarks expose semantic tool schemas in static environments, allowing agents to rely on prior knowledge rather than autonomous discovery. To address this limitation, we introduce ScrambleToolBench, an interactive terminal benchmark designed to isolate behavioral reasoning. By removing semantic cues and enforcing a continuous task curriculum, the benchmark requires agents to uncover hidden tool behaviors entirely through trial-and-error interaction. The benchmark further introduces dynamic challenges, including mapping drift, stochastic action failures, and temporal execution windows, to evaluate whether agents can revise and adapt their hypotheses as the environment changes. Our evaluation of state-of-the-art language models reveals that successful initial discovery does not translate into robust adaptation. When faced with structural changes such as mapping drift, agents fail to use deductive strategies such as cycle tracing, and instead exhibit belief inertia or fall back to exhaustive search. Increasing test-time reasoning only amplifies this expensive brute-force search rather than enabling deductive recovery. While equipping agents with persistent memory reduces compounding errors, they remain unable to efficiently infer structural changes, highlighting a gap in current agent reasoning.

\texttt{Homebot} is a locally deployable AI agent for conversational household assistance and automation. It accepts voice and instant-messaging requests through a shared runtime that combines language-model responses with registered tools and task-specific skills. The design separates common request processing from session ownership: messaging history remains scoped to a channel and chat, whereas voice interaction is bounded by wake-word activation. For hands-free use, \texttt{Homebot} combines local wake-word detection, streaming speech recognition and synthesis, and an explicit dialogue-state protocol for ending, following up, or continuing a conversation. Clear channel, tool, and skill contracts support practical customization for household use.