================================================================
本文档落款 2026-09-29 之前，记载转型前研究；自 2026-09-29 起公司主线唯一为 N=1（运营 Miao 的人生），AI 代采购/AI Data Go 产品线已撤销。
================================================================

================================================================
P0 Internal Test Record (2026-09-25)
P0 内测记录（2026-09-25）
----------------------------------------------------------------
vertciti Data Room · Public Document
Date: 2026-09-27 | Version: v1
Source: internal decision record D1 (2026-09-25),
  deliverables/data-room-numbers.md #8,
  deliverables/oss-landscape.md (pytest 139 passed).
Classification: Public — internal test summary; no private
  keys, credentials, or confidential information.
================================================================

----------------------------------------------------------------
ENGLISH
----------------------------------------------------------------

STATEMENT OF SCOPE (READ FIRST)

One completed run of the internal Python prototype test
suite (P0: the five self-built modules — retry base,
integer-amount money, webhook dispatcher, x402 facilitator
client + settlement state machine, pgvector hybrid
retrieval). The figures below show the codebase passed its
tests at that time. They are NOT a production audit, NOT a
security review, and NOT proof of public-service
availability. The prototype is an in-memory mock.

1. THE NUMBERS (single run, 2026-09-25)

  - pytest: 139 passed · 1 skipped (no local
    Postgres/pgvector available for the DB integration test)
  - End-to-end checks: 70/70 passed
  - x402 regression checks: 14/14 passed
  - Production-live claims: 0 (internal validation only)

2. WHERE EACH NUMBER COMES FROM

  - 139 passed / 1 skipped — internal decision record D1,
    2026-09-25 ("139 tests passed (1 skipped: no local PG)")
  - 70/70 end-to-end — deliverables/data-room-numbers.md #8
    ("prototype phase two: 70 e2e all passed")
  - 14/14 x402 regression — deliverables/data-room-numbers.md
    #8 ("x402 live-link regression 14/14"); the on-chain leg
    of that regression is the 0.01 USDC live-testnet
    settlement documented separately in
    x402-live-testnet-evidence.txt

3. WHAT THIS DOES NOT PROVE

  Does NOT prove: production readiness, mainnet capability,
  revenue, customer activity, or security posture. One
  passing run is evidence the code matched its tests on that
  day — nothing more.

----------------------------------------------------------------
中文
----------------------------------------------------------------

口径声明（先读）

内部 Python 原型测试套件（P0：五个自研模块——重试基座、
金额整数化、webhook 分发器、x402 facilitator 客户端＋结算
状态机、pgvector 混合检索）的一次已完成运行。以下数字只
说明当时代码通过了测试，不是生产审计、不是安全审查、不
代表公网服务可用。原型为内存 mock。

一、数字（2026-09-25 单次运行）

  - pytest：139 通过 · 1 跳过（本地无 Postgres/pgvector，
    DB 集成测试跳过）
  - 端到端检查：70/70 通过
  - x402 回归检查：14/14 通过
  - 生产上线声明：0（仅内部验证）

二、每个数字的来源

  - 139 通过 / 1 跳过——内部决策记录 D1，2026-09-25
    （"139 测试通过（1 跳过：本地无 PG）"）
  - 70/70 端到端——deliverables/data-room-numbers.md #8
    （"原型二期 70 项 e2e 全过"）
  - 14/14 x402 回归——deliverables/data-room-numbers.md #8
    （"x402 真实链路回归 14/14"）；其中链上部分即
    x402-live-testnet-evidence.txt 记录的 0.01 USDC
    live-testnet 结算

三、这不证明什么

  不证明：生产就绪、主网能力、收入、客户行为或安全水位。
  一次通过只说明当天的代码符合当时的测试——仅此而已。

================================================================
END OF DOCUMENT
================================================================