← 研踪 PaperTrailarXiv:2611.05094 · v1
A Gated Spec-Driven Workflow: An Audit-Trail Study of Coding-Agent Work
首次提交 2026-11-03 23:52
版本更新 2026-11-03 23:52
雷达观测 2026-11-04 23:39
摘要
Spec-driven development (SDD) asks a coding agent to work from specifications, plans, and task lists rather than from the chat transcript. We study one such workflow for Claude Code, in which almost every control is an instruction carried out by the same model that it constrains. The user must approve each stage before the next begins. A reviewer subagent checks each artefact against a fixed, stage-specific defect list, and an executor subagent may only touch the files listed in its task. After each block of tasks, the orchestrator runs a file-inventory check to detect undeclared changes. We analyse the logs and transcripts of nine features in seven projects, all run by the authors, ranging from a Fortran solver and Python programs to books and this paper. Of these, five finished with every planned task done, including the Fortran solver feature, and 288 of the 581 planned tasks were completed block by block. The executor kept to its declared files: it left only one undeclared change in the deliverable folder, and the check caught it, although its reports record other departures from the protocol. The weak point was the orchestrator, which runs and judges its own checks. It recorded undeclared changes in 11 of 247 block checks, mostly from the authors' in-round requests and the environment, and stopped work in only four.
作者备注：15 pages, 2 figures, 6 tables
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-03 23:52 · 最新观测版本
A Gated Spec-Driven Workflow: An Audit-Trail Study of Coding-Agent Work
Spec-driven development (SDD) asks a coding agent to work from specifications, plans, and task lists rather than from the chat transcript. We study one such workflow for Claude Code, in which almost every control is an instruction carried out by the same model that it constrains. The user must approve each stage before the next begins. A reviewer subagent checks each artefact against a fixed, stage-specific defect list, and an executor subagent may only touch the files listed in its task. After each block of tasks, the orchestrator runs a file-inventory check to detect undeclared changes. We analyse the logs and transcripts of nine features in seven projects, all run by the authors, ranging from a Fortran solver and Python programs to books and this paper. Of these, five finished with every planned task done, including the Fortran solver feature, and 288 of the 581 planned tasks were completed block by block. The executor kept to its declared files: it left only one undeclared change in the deliverable folder, and the check caught it, although its reports record other departures from the protocol. The weak point was the orchestrator, which runs and judges its own checks. It recorded undeclared changes in 11 of 247 block checks, mostly from the authors' in-round requests and the environment, and stopped work in only four.
查看此版本解读 →查看此版本原文 ↗