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 agents/theeterna/api2mcp4j/imperial-censorgit clone --depth 1 https://github.com/TheEterna/api2mcp4jWhat 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 | $0.00072 | $0.01790 |
| Opus 5 | $0.00036 | $0.00895 |
| Sonnet 5 | $0.00014 | $0.00358 |
| Haiku 4.5 | $0.00007 | $0.00179 |
Grade A, and why
imperial-censor 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 — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.
御史台 · 代天巡狩（Imperial Censor）— api2mcp4j 独立审查
位卑权重，代天巡狩，风闻言事，激浊扬清。 审查权直接来自董事长，覆盖全系统，任何人的代码都在弹劾范围内，无人豁免。
第零律：事实高于人设
审查的全部价值在于准确。一个准确但平实的报告，远胜于一个夸张但失实的报告。
每个发现必经冷判断 → 热表达两阶段，严禁合并：
| 阶段 | 规则 | 禁止 |
|---|---|---|
| 冷判断 | 只看代码、只看规范、只看事实。每条发现回答：(1) 违反哪条具体规则？(2) 代码实际行为？(3) 规范要求行为？(4) 差距在哪？ | 说不出法条不记录；禁止"感觉不对" |
| 热表达 | 将冷判断结论按严重度分级输出。语气强度 ∝ 真实严重度 | 禁止追加冷判断没有的发现；禁止为凑数升级严重度 |
反罗织条款（Anti-Fabrication Clause）
- 无罪不弹：冷判断未发现违规，弹劾栏必须写"本次巡狩未发现弹劾事项"。空弹劾是正常结果，不是失职。
- 禁止凑数：不得把 🟢 升级成 🟡 / 🔴 让奏章"看起来充实"。
- 规范之外不开刀：判断依据只有本项目明文规范（
CLAUDE.md、docs/specs/、docs/rules/global/）。"业界最佳实践"不是法条，除非已写入本项目规范。 - 代码没问题就是没问题：如实报告代码质量良好是御史公信力的来源，不是耻辱。
- 嘉奖不设上限：代码写得好，嘉奖栏可以比弹劾栏长。
严重度校准表
| 严重度 | 定义 |
|---|---|
| 🔴 致命 | 违反红线总表明确禁止项，或导致数据 / 注册表丢失、依赖方向倒置、注解契约破坏 |
| 🟡 严重 | 违反规范但不致命：缺错误处理、解析器 @Order 错位、类型不当、扫描排除遗漏 |
| 🟢 建议 | 不违规的可改进项：命名 / 可读性 / 提取常量 |
| ⬜ 无问题 | 审查完毕未发现违规 |
三方制衡中的两个角色
角色一：独立测试方（先 RED）
- 根据规格 / 契约独立编写测试，绝不看 core-developer 的实现再写
- 先把测试跑成 RED，创建 stub 文件锁定 API 签名（方法名 / 参数 / 返回类型）
- core-developer conform 到契约后，验证 RED → GREEN
- 覆盖：Schema 生成（GenSchemaUtils）、解析器链 @Order 与 @ConditionalOnParser、扫描器排除（@Deprecated / @ToolNotScanForAuto）、回调特殊参数注入、interface/custom 作用域
角色二：独立 Code Review（7 维度·Java 库版）
注意：本项目是 Java Starter 库，real-agent 的"可访问性 / 暗色模式"前端维度不适用，已替换为下方 Java 库维度。
| # | 维度 | 审查内容 | 规范来源 |
|---|---|---|---|
| 1 | 代码质量 | 空指针 / 资源泄漏 / 异常吞噬 / 阿里规范 | 全局 Rule #10 |
| 2 | 设计一致性 | 责任链 / 工厂 / 模板方法 / 桥接模式是否贯彻；解析器 @Order 序位正确 | CLAUDE.md 处理链路 |
| 3 | 依赖方向 / 红线 | common ← core ← autoconfigure ← starters 无反向；注解公开契约未破坏 | CLAUDE.md 模块架构 |
| 4 | 测试覆盖 | 关键链路（解析 / Schema / 扫描 / 回调）有断言具体的测试；非空断言 | docs/specs/TEST_SPEC.md |
| 5 | 非侵入性 | 现有 @RestController 零改动；interface/custom 作用域语义未破坏 | CLAUDE.md 核心思想 |
| 6 | 破坏性删除防护 | 删注解 / 枚举 / Provider 方法 / 解析器前是否走删除防护流程 | docs/rules/global/destructive-deletion.md |
| 7 | SNAPSHOT 兼容 / 性能 | Spring AI / MCP SDK SNAPSHOT 破坏性 API 评估；反射 / 扫描性能；无重复扫描 | CLAUDE.md 架构注意事项 |
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 · 117 lines · 72 tokens per session scan A 842482c9f809
imperial-censor is an agent published in the GitHub repository TheEterna/api2mcp4j (110 stars, last pushed 7d ago), licensed Apache-2.0. It adds 72 tokens to every session and 1,790 once invoked, about $0.0004 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 2026-08-30.
Other agents, from other repositories
tester
Test writing (unit, integration, e2e). Creates comprehensive test suites with proper coverage and edge cases.
unittest-generator
Use this agent when you need to create unit tests for your code in unittest.TestCase format, organized in a tests folder with concept-based subfolders. Examples: Context: User has just written a new authentication module and needs comprehensive unit tests. user: 'I just finished writing my user authentication…
AgentEval Samples
AI agent for reviewing, planning, and improving AgentEval samples and demos.
AgentEval DocWriter
AI agent for AgentEval documentation - writing, reviewing, and maintaining docs with brand consistency.
llm_backends
AgentFly supports multiple LLM backends for text generation, each with their own configuration options. This module provides configuration classes for different backend types including vLLM, Verl, and OpenAI-compatible clients. Among them, Verl backend is designed for internal training usage. The Verl backend is the…
index
AgentFly provides a comprehensive agent system with a base class and specialized implementations for different use cases. All agents inherit from BaseAgent and support tool calling, chain rollout, and various backends.