When Only the Final Text Survives: Implicit Execution Tracing for Multi-Agent Attribution

TL;DR

This paper introduces Implicit Execution Tracing (IET), a novel framework that embeds agent-specific signals into generated text to enable accountability and attribution even when execution logs are unavailable.

Contribution

IET shifts attribution from post-hoc inference to built-in instrumentation, embedding provenance directly into text for robust multi-agent attribution without relying on execution metadata.

Findings

IET achieves accurate segment-level attribution in diverse settings.

IET reliably recovers execution traces under identity removal and privacy redaction.

IET maintains generation quality while embedding provenance signals.

Abstract

When a multi-agent system produces an incorrect or harmful answer, who is accountable if execution logs and agent identifiers are unavailable? In practice, generated content is often detached from its execution environment due to privacy or system boundaries, leaving the final text as the only auditable artifact. Existing attribution methods rely on full execution traces and thus become ineffective in such metadata-deprived settings. We propose Implicit Execution Tracing (IET), a provenance-by-design framework that shifts attribution from post-hoc inference to built-in instrumentation. Instead of reconstructing hidden trajectories, IET embeds agent-specific, key-conditioned statistical signals directly into the token generation process, transforming the output text into a self-verifying execution record. At inference time, we recover a linearized execution trace from the final text via transition-aware statistical scoring. Experiments across diverse multi-agent coordination settings demonstrate that IET achieves accurate segment-level attribution and reliable transition recovery under identity removal, boundary corruption, and privacy-preserving redaction, while maintaining generation quality. These results show that embedding provenance into generation provides a practical and robust foundation for accountability in multi-agent language systems when execution metadata is unavailable.