View-oriented Conversation Compiler for Agent Trace Analysis
Lvmin Zhang, Maneesh Agrawala

TL;DR
This paper introduces VCC, a compiler that transforms complex agent JSON logs into structured views, improving analysis quality and efficiency in agent trace analysis.
Contribution
VCC provides a novel compilation approach that creates multiple structured views from raw agent logs, enhancing analysis and context engineering capabilities.
Findings
Replacing raw JSONL with VCC views increases model pass rates.
VCC reduces reflector token consumption by 50-66%.
Structured views lead to more concise learned memory.
Abstract
Agent traces carry increasing analytical value in agentic systems and context engineering, yet most prior work treats conversation format as a trivial implementation detail. Modern agent conversations, however, contain deeply structured content, including nested tool calls and results, chain-of-thought reasoning blocks, sub-agent invocations, context-window compaction boundaries, and harness-injected system directives, whose complexity far exceeds that of simple user-assistant exchanges. Feeding such traces to a reflector or other analytical mechanism in plain text, JSON, YAML, or via grep can materially degrade analysis quality. This paper presents VCC (View-oriented Conversation Compiler), a compiler (lex, parse, IR, lower, emit) that transforms raw agent JSONL logs into a family of structured views: a full view (lossless transcript serving as the canonical line-number coordinate…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
