arXiv cs.MA· Lei Zhenbing, Lei Chang·· 11 小时前AI 评分34
研究：多智能体审计层中 10 条自我声称的修复声明经机械复验，仅半数在指定位置真实存在
PASS Is a Claim, Not an Observation: Mechanical Re-verification of 10 Self-Asserted Fix Claims in a Live Multi-Agent Audit Layer
AI 导读
研究对 AITRAP 多智能体知识树平台中协调智能体 azone_god 的 10 条自我声称代码修复声明做了机械复验，这些声明提取自 179 条协调广播。三项检查通过率分别为文件修改时间一致性 70%、指定标识符与调用点存在防御模式 60%、仅凭广播内容规则化复验 80%。关键发现是：在信息足够复验的声明中，50% 未通过标识符+调用点检查，修复在声称位置或整个代码库中并不存在。
正文
Abstract:When agents claim to have fixed their own defects, how do we know the fix is real? We mechanically re-verified 10 self-asserted code-level fix claims from a privileged coordination agent (azone_god) operating in AITRAP, a live multi-agent knowledge-tree platform. These 10 claims were extracted from a corpus of 179 coordination broadcasts (28 containing structured fix assertions, 31 unique fix-ID/message pairs, 18 distinct fix identifiers). Fix claims use mixed English/Chinese formats (PASS, checkmarks, and "verification passed" declarations). Each code-level claim was checked against three criteria: (1) file modification timestamp (mtime) consistency -- 70% pass; (2) presence of the claimed defensive pattern at the specified identifier and call site -- 60% pass in the claimed file (80% anywhere in the codebase); (3) deterministic rule-based re-verification from the broadcast announcement alone -- 80% pass. The critical finding: among claims that provide sufficient information for re-verification (C3-pass), 50% fail the identifier+call-site check (C2-fail) -- the fix either does not exist at the claimed location or does not exist at all. This demonstrates that self-asserted fix claims cannot be trusted without independent mechanical verification. Our central claim: a PASS declaration is a claim requiring independent verification, not an observation of fix completion.
| Comments: | 8 pages, 4 tables. v1.1 revised Oct 24, 2026 |
| Subjects: | Multiagent Systems (cs.MA); Software Engineering (cs.SE) |
| Cite as: | arXiv:2610.35950 [cs.MA] |
| (or arXiv:2610.35950v1 [cs.MA] for this version) | |
| https://doi.org/10.48550/arXiv.2610.35950 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Lei Zhenbing [view email]
[v1]
Sat, 24 Oct 2026 12:29:26 UTC (10 KB)
来源：arXiv cs.MA · arxiv.org