When Errors Become Narratives: A Longitudinal Taxonomy of Silent Failures in a Production LLM Agent Runtime

• URL: http://arxiv.org/abs/2606.14589v1
• Date: Fri, 12 Jun 2026 16:06:55 GMT
• Title: When Errors Become Narratives: A Longitudinal Taxonomy of Silent Failures in a Production LLM Agent Runtime
• Abstract summary: We present a longitudinal study of silent failures in a personal-assistant agent runtime in continuous production since March 2026.<n>We documented 22 incidents with full root-cause postmortems, in which one meta-pattern manifested at least 28 times.<n>We derive a five-class, mechanism-oriented taxonomy: (A) environment and platform quirks, (B) design-assumption mismatches, (C) error swallowing and dilution, (D) chained hallucination and fabrication, (E) operational omission and forensic blind spots.
• Score: 4.56904471801595
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: LLM agent systems increasingly run as long-lived autonomous runtimes: scheduling jobs, calling tools, maintaining memory, and pushing results to humans. We present a longitudinal study of silent failures in one such system: a personal-assistant agent runtime in continuous production since March 2026, with roughly 40 scheduled jobs, 8 LLM providers, a tool-governance proxy, and a knowledge-base memory plane, defended by 4,286 unit tests and 827 governance checks. Over eight weeks we documented 22 incidents with full root-cause postmortems, in which one meta-pattern -- a failure whose error signal never reaches a human in actionable form -- manifested at least 28 times. We derive a five-class, mechanism-oriented taxonomy: (A) environment and platform quirks, (B) design-assumption mismatches, (C) error swallowing and dilution, (D) chained hallucination and fabrication, (E) operational omission and forensic blind spots. Class D is unique to LLM systems and the most dangerous: the system does not merely fail to report an error -- the LLM transforms it into fluent, plausible narrative delivered to the user. We term this fail-plausible: gray failure's differential observability escalated -- the observer is not just blind, it is convincingly lied to by the failure itself. Three findings: about 70% of silent failures were caught by human user-view observation, not tests or audits; a retrospective audit of 15 incidents found 0% ex-ante prevention but 87% regression blocking -- audits are regression engines, not prediction engines; incident latency (13 hours to 60 days) tracks failure mechanism, not code complexity -- the longest-lived failures lived in the seams between components, where no test runs. We describe the resulting defense framework and distill design principles for agent systems whose failures are loud, attributable, and boring. All postmortems and artifacts are public.

Related papers

• Why LLM Agents Collapse Without Oversight: The Enforcement Gap as the Mechanism Behind Emergence World Failures [19.162716475205308]
  Reflexion-style agents already detect dangerous plan steps through iterative self-critique.<n>We call this the enforcement gap: the audit sees the problem; the controller ignores it.<n>We prove formally that when enforcement probability is near zero, detection quality is irrelevant to security.
  arXiv  Detail & Related papers  (2026-09-14T09:46:59Z)
• DCFA: Dual-view Causal-inspired Attribution for Failure Reasoning in LLM-based Multi-agent Systems [61.11435469308521]
  We propose DCFA, a training-free framework for failure attribution in large language model systems.<n> DCFA integrates a global module that constructs structured causal-inspired dependency graphs from system traces.<n>Experiments show that DCFA improves step-level accuracy by up to 8.27% over state-of-the-art baselines.
  arXiv  Detail & Related papers  (2026-09-04T05:29:31Z)
• Agentic Security: A Systematization of Tools, Failure Modes, and Design Laws for LLM-Driven Penetration Testing [0.21990652930491858]
  Agentic security uses large-language-model (LLM) agents to plan, dispatch, and interpret security tools.<n>We systematize these failures through a hands-on evaluation of ten widely used static, dynamic, cloud, orchestration, and AI red-teaming tools.
  arXiv  Detail & Related papers  (2026-08-16T07:18:04Z)
• SearchAuditor: Auditing and Attributing Failures in Long-Horizon Search Agents [56.400332595452966]
  We introduce SearchAuditBench, a benchmark that evaluates whether auditors can localize, attribute, and repair search-agent failures.<n>SearchAuditBench comprises 1,243 failed trajectories, averaging 73.1 messages and 65.1K tokens, collected from eight open-weight models on five deep-search benchmarks.<n>We propose SearchAuditor, a multi-perspective auditing framework that effectively localizes, attributes, and repairs search-agent failures.
  arXiv  Detail & Related papers  (2026-08-05T09:09:00Z)
• Evaluation Blindness: How Silent Measurement Failures Corrupt AI Systems from Training to Deployment [0.0]
  The failure propagates through training loops, evaluation pipelines, and production monitoring stacks until downstream harm makes it visible.<n>This paper introduces evaluation blindness: a measurement function M exhibits evaluation blindness with respect to failure class F when it produces readings indistinguishable from a healthy state while the system is actually failing.
  arXiv  Detail & Related papers  (2026-08-03T18:33:37Z)
• Fantastic Adaptive Taxonomies and How to Use Them [100.20614173157708]
  AdaMAST converts execution traces into a compact, evidence-grounded failure taxonomy.<n>No code is hand-authored, no trace human-annotated.<n>In agent-system search, taxonomy-coded diagnoses of failed candidates outperform free-form.
  arXiv  Detail & Related papers  (2026-07-17T17:41:16Z)
• SIMMER: Benchmarking Latent Failures in LLM Executable Planning with a World Model [8.782590561222511]
  We introduce SIMMER, a benchmark for evaluating latent failures in large language models (LLMs) planning.<n> SIMMER defines a world model comprising 77 actions, 262 unique objects, and approximately 46,800 possible interactions that are semantically realistic.<n>It then leverages a state machine executor that validates plans against the world model and detects immediate precondition violations, latent hazards, and irreversible failures.
  arXiv  Detail & Related papers  (2026-06-12T15:53:16Z)
• Catching One in Five: LLM-as-Judge Blind Spots in Production Multi-Turn Transaction Agents [45.148328075418156]
  We study a deployed multi-turn food-and-beverage ordering agent and measure how many genuine quality problems it catches.<n>Our blind-spot taxonomy shows the failure is structured, not random.<n>For production multi-turn agents, automated judging is a regression floor, not a substitute for human review.
  arXiv  Detail & Related papers  (2026-06-09T02:11:01Z)
• Characterizing the Failure Modes of LLMs in Resolving Real-World GitHub Issues [10.34693262597204]
  Large Language Models (LLMs) are increasingly deployed to resolve real-world GitHub issues.<n>This paper evaluates three state-of-the-art models, i.e., Claude 4.5 Sonnet, Gemini 3 Pro, and GPT-5, on the SWE-bench Verified dataset.<n>We conduct a rigorous manual analysis of the symptoms and root causes underlying 243 failed attempts across 900 total trials.
  arXiv  Detail & Related papers  (2026-05-12T15:34:57Z)
• FLARE: Agentic Coverage-Guided Fuzzing for LLM-Based Multi-Agent Systems [9.562231408217281]
  Multi-Agent LLM Systems (MAS) have been adopted to automate complex human tasks by breaking down tasks into subtasks.<n>MAS applications frequently encounter failures, including infinite loops and failed tool invocations.<n>Traditional software testing techniques are ineffective in detecting such failures due to the lack of LLM agent specification, the large behavioral space of MAS, and semantic-based correctness judgment.<n>This paper presents FLARE, a novel testing framework tailored for MAS. FLARE takes the source code of MAS as input and extracts specifications and behavioral spaces from agent definitions.
  arXiv  Detail & Related papers  (2026-04-07T00:47:37Z)
• Evaluating & Reducing Deceptive Dialogue From Language Models with Multi-turn RL [64.3268313484078]
  Large Language Models (LLMs) interact with millions of people worldwide in applications such as customer support, education and healthcare.<n>Their ability to produce deceptive outputs, whether intentionally or inadvertently, poses significant safety concerns.<n>We investigate the extent to which LLMs engage in deception within dialogue, and propose the belief misalignment metric to quantify deception.
  arXiv  Detail & Related papers  (2025-10-16T05:29:36Z)
• AgenTracer: Who Is Inducing Failure in the LLM Agentic Systems? [71.21547572568655]
  AgenTracer-8B is a lightweight failure tracer trained with multi-granular reinforcement learning.<n>On the Who&When benchmark, AgenTracer-8B outperforms giant proprietary LLMs like Gemini-2.5-Pro and Claude-4-Sonnet by up to 18.18%.<n>AgenTracer-8B delivers actionable feedback to off-the-shelf multi-agent systems like MetaGPT and MaAS with 4.8-14.2% performance gains.
  arXiv  Detail & Related papers  (2025-09-03T13:42:14Z)
• Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems [50.29939179830491]
  Failure attribution in LLM multi-agent systems remains underexplored and labor-intensive.<n>We develop and evaluate three automated failure attribution methods, summarizing their corresponding pros and cons.<n>The best method achieves 53.5% accuracy in identifying failure-responsible agents but only 14.2% in pinpointing failure steps.
  arXiv  Detail & Related papers  (2025-04-30T23:09:44Z)
• Discovering Knowledge Deficiencies of Language Models on Massive Knowledge Base [30.705524808195268]
  Large language models (LLMs) possess impressive linguistic capabilities but often fail to faithfully retain factual knowledge.<n>We propose error ascent (SEA), a scalable and efficient framework for discovering knowledge deficiencies (errors) in closed-weight LLMs.<n>SEA uncovers 40.7x more knowledge errors than Automated Capability Discovery and 26.7% more than AutoBencher.
  arXiv  Detail & Related papers  (2025-03-30T08:33:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.