Observability for Delegated Execution in Agentic AI Systems
2026-06-08 • Cryptography and Security
Cryptography and SecurityArtificial Intelligence
AI summaryⓘ
The authors explain that it's hard to figure out which parts of a task were done by whom when many helpers (agents) work together, especially with AI systems that use multiple tools and create sub-agents. Usual methods like logs and traces look the same even if the actual assignments differ, making it confusing to track responsibility. They propose a new system that keeps track of who did what in real-time, allowing accurate reconstruction across different tools without guessing based on time. Their focus is on tracing actions under delegation, not on figuring out intentions or thoughts behind actions.
delegation-scoped executionaudit logsexecution tracesLLM-based agentssub-agentscausal reconstructionobservability substrateinformation modelforensic queriescross-tool attribution
Authors
Abhinav Mishra, Kumar Sharad
Abstract
Delegation-scoped execution is not identifiable from standard observables: audit logs and execution traces can be identical under multiple incompatible delegation assignments. This gap is especially acute in LLM-based agentic systems, where agents dynamically select tools, vary execution sequences across runs for the same instruction, and spawn cooperating sub-agents. These dynamics fragment and interleave traces, making delegation-scoped reconstruction from causal structure alone structurally underdetermined. Although individual actions are authorized and logged, existing audit, tracing, and security schemas lack the semantics to reconstruct what actions occurred under a given delegation across heterogeneous systems. We focus on delegation-scoped attribution and access/share footprint reconstruction, not intent inference or reasoning reconstruction. We present an agent-aware observability substrate consisting of a lightweight gateway and a common information model that binds delegation context at execution time. This enables reliable cross-tool delegation-scoped reconstruction and direct forensic queries without heuristic time-window correlation.