ObjectGraph: From Document Injection to Knowledge Traversal -- A Native File Format for the Agentic Era

TL;DR

OBJECTGRAPH is a new file format that represents documents as knowledge graphs, enabling more efficient and effective interaction with autonomous LLM agents by reducing token usage and preserving content fidelity.

Contribution

The paper introduces OBJECTGRAPH, a novel document format that models documents as knowledge graphs, addressing structural limitations of existing formats for agent consumption.

Findings

Up to 95.3% token reduction in agent tasks.

No significant degradation in task accuracy (p > 0.05).

98.7% content preservation in transpiler benchmark.

Abstract

Every document format in existence was designed for a human reader moving linearly through text. Autonomous LLM agents do not read - they retrieve. This fundamental mismatch forces agents to inject entire documents into their context window, wasting tokens on irrelevant content, compounding state across multi-turn loops, and broadcasting information indiscriminately across agent roles. We argue this is not a prompt engineering problem, not a retrieval problem, and not a compression problem: it is a format problem. We introduce OBJECTGRAPH (.og), a file format that reconceives the document as a typed, directed knowledge graph to be traversed rather than a string to be injected. OBJECTGRAPH is a strict superset of Markdown - every .md file is a valid .og file - requires no infrastructure beyond a two-primitive query protocol, and is readable by both humans and agents without tooling. We formalize the Document Consumption Problem, characterise six structural properties no existing format satisfies simultaneously, and prove OBJECTGRAPH satisfies all six. We further introduce the Progressive Disclosure Model, the Role-Scoped Access Protocol, and Executable Assertion Nodes as native format primitives. Empirical evaluation across five document classes and eight agent task types demonstrates up to 95.3 percent token reduction with no statistically significant degradation in task accuracy (p > 0.05). Transpiler fidelity reaches 98.7 percent content preservation on a held-out document benchmark.