Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add rules/silenceinsect/aidocxworkflow/stage_s5_test_pointsgit clone --depth 1 https://github.com/SilenceInsect/AIDocxWorkFlowWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00020 | $0.08895 |
| Opus 5 | $0.00010 | $0.04448 |
| Sonnet 5 | $0.00004 | $0.01779 |
| Haiku 4.5 | $0.00002 | $0.00890 |
Grade A, and why
STAGE_S5_TEST_POINTS scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 651 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AIDocxWorkFlow S5 — 测试点生成
独立阶段:可以单独调用。上游材料(S2 backlog)审查合格后开始,失败写失败报告。
第一原则
S5 当前不是在做代码实现,也不是在写最短答案,而是在做需求覆盖设计。 默认目标不是少写,而是找全;不是压缩,而是显式列出覆盖和未覆盖。 任何未设计成 TP 的需求点,不允许静默消失,必须进入
omission_ledger.json。
阶段入口
触发方式:/aidocx-s5-test-points 或粘贴 S2 backlog
前置材料(必须全部满足):
| 材料 | 来源 | 审查要求 |
|---|---|---|
| S2 backlog.md(必须读取) | workflow_assets/<req_name>/<version>/「S2 需求拆解」/backlog.md |
存在且含 Epic/Story |
| S4 business_flow.md(必须读取) | workflow_assets/<req_name>/<version>/「S4 流程图导出」/business_flow.md |
包含异常/错误决策树和风险点清单,用于补充 EXCEPTION 类型测试点 |
前置材料缺失时:生成失败报告,停止 S5。
§1.9 命名一致性 & 产出标准化门禁(字段溯源版)
**(FIELD-001):本节解决 TP 文本语义与 S2 正式命名体系脱节的问题。 校验精度:字段溯源版(字段精准匹配——obj_name/fp_name == S2)+ LLM 自由文本溯源(title/description 不带锚点)
§1.9.1 核心规范
S5 产出的每条测试点必须满足以下要求:
v31 修订:
obj_name来源 =S2 backlog.epics[].title(Epic 标题 = 需求对象);fp_name来源 =S2 backlog.stories[].title(Story 标题 = 功能描述)。
-
显式字段:
obj_name和fp_name字段必须存在obj_name:从 S2backlog.epics[].title取值,100% 原样写入fp_name:从 S2backlog.stories[].title取值,100% 原样写入
-
文本溯源分离:
title:纯场景简短标题(不带锚点),4-12 字场景摘要description:纯测试逻辑(前置+步骤+预期),干净通顺
-
字段精准匹配:
obj_name== S2requirement_objects[].obj_name(逐字相等)fp_name不重复 S2requirement_objects[].feature_points[].fp_desc(LLM 自创中性名)
-
锚点取消:title / description / test_scenario 不再以锚点开头——锚点仅存 JSON 字段,不重复进文本
§1.9.2 校验方式
L1 格式校验层执行,检查项:field_traceability
| 校验项 | 要求 |
|---|---|
obj_name 字段 |
存在且 == S2 obj_name(逐字相等) |
fp_name 字段 |
存在且 LLM 自创(命名规则见 §1.9.1) |
feature_point_ref 字段 |
强制必填——必须引用 S2 OBJ.feature_points[].id |
title 文本 |
不带锚点(无【】) |
description 文本 |
不带锚点(无【】) |
通过率硬门禁:100%(0 错误才可进入 S6)
§1.9.3 不通过处理
- 自动触发 S5 迭代补全,最多重试 3 次
- 3 次仍不通过 →
L1_FAILED状态 → 人工审查队列 - 禁止未通过校验的 TP 进入 S6
§1.9.4 命名来源约束
obj_name:唯一来源 = S2requirement_objects.json的obj_name字段。100% 逐字相等。必须是实体名词。fp_name:唯一来源 = LLM 自创(中性功能名)。约束:必须是动词/动作 + 长度 ≤ 20 字符 + 不与 S2fp_desc字面量重复;禁止名词性描述(如"道具数据配置"/"价格配置"应改为"道具从配置加载"/"道具按配置价格展示")。功能描述字段(JSON 中的功能描述键):必须动作化——来自 S2fp_desc做语义等价改写,将名词性描述改写为动词化表达。- 改写规则:不得改变 S2
fp_desc的事实范围,仅将静态描述动作化。 - 改写模式(参考):
S2 fp_desc(原始) 动作化后(生成 TP 时使用) 配置道具分类归属 按配置对道具分类 配置道具基础信息 按配置表加载道具数据 配置道具价格 按配置校验道具价格 - 改写后同步更新
s2_source.fp_desc字段(记录改写后语义,供溯源)。 s2_source块记录的是实际使用的语义(改写后),不是 S2 原始文本的复述。
- 改写规则:不得改变 S2
- 禁止 LLM 自创
obj_name(必须 S2 严格匹配)。 - 禁止
fp_name与 S2fp_desc字面量重复(避免语义重定义)。
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 651 lines · 20 tokens per session scan A 747f800ea766
STAGE_S5_TEST_POINTS is a cursor rule published in the GitHub repository SilenceInsect/AIDocxWorkFlow (2 stars, last pushed 1mo ago), licensed MIT. It adds 20 tokens to every session and 8,895 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
cli-error-handling
CLI command error handling patterns.
prefer-direct-imports-over-module-mocks
Prefer extracting a testable core over vi.mock / vi.resetModules when unit tests need to reach production logic entangled with config, env, or singletons.
control-plane-descriptors
Control plane descriptor and instance implementation patterns.
family-instance-domain-actions
Family instance domain action implementation patterns.