Summary
Published on arXiv in March 2026 (arXiv:2603.16021), this paper by Jake Van Clief and David McDermott introduces Interpretable Context Methodology (ICM) — the discipline of using filesystem layout as the orchestration mechanism for a multi-step LLM workflow, in place of a code-level multi-agent framework. Its load-bearing claim: if the prompts and context for each stage already exist as files in a well-organized folder hierarchy, you don’t need a coordination framework — you need one agent that reads the right files at the right moment.
Why it matters
The paper applies fifty-year-old engineering ideas — Unix pipelines, one-job-per-stage, plain text as the universal interface — to a contemporary problem: delivering scoped context to a model across stages without a monolithic prompt. It is the canonical source for Interpretable Context Methodology, and a close cousin of the agent-readable web standards: both are file-native context delivery. Where AGENTS.md asks how a repository tells an agent what to do, ICM asks how a workspace does. This is a copyrighted academic paper; only an original summary and the canonical link are kept here.
Part of the AP0110.ORG source library. See the wiki overview.
Full source text: read the original document.