Formal Architecture Descriptors as Navigation Primitives for AI Coding Agents
Ruoqi Jin

TL;DR
Providing AI coding agents with formal architecture descriptors significantly reduces navigation overhead, improves accuracy, and enhances consistency in code exploration tasks.
Contribution
The paper introduces intent.lisp, an S-expression architecture descriptor, and demonstrates its effectiveness through experiments and a field study, advancing AI code navigation methods.
Findings
Architecture context reduces navigation steps by 33-44%.
Automatically generated descriptors achieve 100% accuracy.
52% reduction in behavioral variance in field sessions.
Abstract
AI coding agents spend a substantial fraction of their tool calls on undirected codebase exploration. We investigate whether providing agents with formal architecture descriptors can reduce this navigational overhead. We present three complementary studies. First, a controlled experiment (24 code localization tasks x 4 conditions, Claude Sonnet 4.6, temperature=0) demonstrates that architecture context reduces navigation steps by 33-44% (Wilcoxon p=0.009, Cohen's d=0.92), with no significant format difference detected across S-expression, JSON, YAML, and Markdown. Second, an artifact-vs-process experiment (15 tasks x 3 conditions) demonstrates that an automatically generated descriptor achieves 100% accuracy versus 80% blind (p=0.002, d=1.04), proving direct navigational value independent of developer self-clarification. Third, an observational field study across 7,012 Claude Code…
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.
