AuditCoder: Responsibility-Preserving Task Graphs for Auditable Code Generation and Bounded Repair
Source: https://arxiv.org/abs/2607.29529
Authors: Kangjie Huang, Chen Lyu
Published: 2026-07-31
Categories: cs.SE
Grounding: opencli arxiv metadata + Obsidian paper digest excerpt (OpenClaw定时任务/论文流水线/2026-08-04-论文流水线.md)
摘要（中文）
论文把代码生成结果和可审计构造轨迹一起生成AuditCoder 在生成前构造带 contract 的任务图，把责任身份实现来源验证证据和修复历史绑定到节点验证失败时只定位并重生成对应区域，未受影响部分保持冻结摘要报告 APPS pass@1 为 82.583.0%，ClassEval 为 75.082.0%；另有 200 条 APPS 记录审计，task-macro decision-code trace coverage 为 0.9725
Abstract
Code generators return programs, but typically do not preserve the construction record needed to connect a failure to the decision that produced the affected code or to delimit a justified repair. We present AuditCoder, which treats the program and an auditable construction trace as joint outputs. Before code generation, a contract-annotated task graph assigns stable responsibility identities that remain attached to each commitment, its owned implementation, provenance, validation evidence, and intervention history. When validation fails, a conservative locator maps heterogeneous evidence to a node or dependency branch---or abstains---and bounded repair regenerates only that region while reusing the frozen complement.
Intake rationale
- Category: coding
- Quality score: 4/5
- Why it matters: AuditCoder binds generated code to task-graph responsibility and bounded repair evidence.
Obsidian evidence excerpt
论文把代码生成结果和可审计构造轨迹一起生成。AuditCoder 在生成前构造带 contract 的任务图，把责任身份、实现、来源、验证证据和修复历史绑定到节点。验证失败时只定位并重生成对应区域，未受影响部分保持冻结。摘要报告 APPS pass@1 为 82.5–83.0%，ClassEval 为 75.0–82.0%；另有 200 条 APPS 记录审计，task-macro decision-code trace coverage 为 0.9725。