tribunal
Skill0 starsChanged 2 months agoMITView on GitHub
对已成形的产品、UI 或代码改动做证据驱动的多视角对抗审查,交叉验证真实用户、设计与实现风险,并在用户授权时修复最高价值问题。用于主要功能、大改或里程碑的深审;不用于微小改动、纯审美发散、尚无可运行产物的空谈,也不把单纯 review 自动扩大成修改或部署。
12 other projects use a version of this.
Adopted by
Most recently taken up by
- a-tokyo/agent-skillstheir own version, first seen Nov 8, 2026
- SpecialITWeapons/tribunal-claude-skilltheir own version, first seen Nov 8, 2026
- Hive-Academy/PtahCodetheir own version, first seen Nov 8, 2026
- Hive-Academy/ptah-extensiontheir own version, first seen Nov 3, 2026
- Tcuzzo/backs-aios-skillstheir own version, first seen Nov 3, 2026
- team-attention/hoyeontheir own version, first seen Oct 8, 2026
Used with
Projects with it also use production-grade, session-handoff and 1 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
Sean-MichaelSkillproduction-grade
Apply production-grade standards when building or modifying web applications. Trigger on: "make it production ready", "add auth", "production grade", "harden this", or any time a user is building a web app with user accounts, an API, or a database and hasn't said it's a prototype. Standards are the default — not an add-on.
Also in 3 projects that use this
wpankSkillsession-handoff
WHAT: Create comprehensive handoff documents that enable fresh AI agents to seamlessly continue work with zero ambiguity. Solves long-running agent context exhaustion problem.
Also in 3 projects that use this
SentrySkillskill-creator
Create new agent skills following the Agent Skills specification. Use when asked to "create a skill", "add a new skill", "write a skill", "make a skill", "build a skill", or scaffold a new skill with SKILL.md. Guides through requirements, writing, registration, and verification.
Also in 3 projects that use this
Other versions12 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 12
- team-attention/hoyeon★ 173same name
- Hive-Academy/ptah-extension★ 34same name
- Hive-Academy/PtahCode★ 34same name
- a-tokyo/aiworkspace★ 19same name
- a-tokyo/agent-skills★ 18same name
- a-tokyo/agent-skills-harness★ 11same name
- cukas/claudes-ai-buddies★ 9same name
- juliajakubowskabiznes/claude-skills★ 2same name
- Tcuzzo/backs-aios-skills★ 2same name
- hekman316/claude-skill-tribunal★ 1same name
2 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
Files
Tribunal
Tribunal 的价值是让不同视角攻击同一个真实产物,再由主 agent 用证据裁决。它不是 critique 数量比赛,也不是固定四人仪式。
先判授权
- 用户只说“审查、挑刺、反驳、评估”时,默认 review-only:只读取证并给结论,不改文件、不部署、不提交。
- 用户明确说“审查并修复、找完直接改”时,进入 review-and-fix。
- 项目契约已经授权的动作仍按项目契约;Tribunal 本身不扩张授权。
启动条件
适用:已有可运行产物或可检查 diff 的主要功能、大改、里程碑、上线前高风险路径。
不适用:
- 改一个字号、文案或小函数;
- 只有方向选择、还没有产物;
- 用户明确要求快速定向检查;
- 没有任何可复现证据,却想靠角色扮演凑意见。
默认预算
- 选择 1 条最高价值主链路和与本次改动直接相关的边界状态。
- 最多 3 个独立审查视角;并发槽不足就合并角色,不为凑人数新建用户 task/thread。
- 每个视角最多提交 5 条发现;主 agent 最终最多接纳 3 个 P0/P1 问题。
- 用户明确要求 exhaustive review 时才扩大覆盖面,并先服从用户给定的时间或成本上限。
工作流
1. 建立共同物证
主 agent 先完成最小基线:
- 读取项目契约、工作区状态和本次 diff,保护用户已有改动。
- 确认真实入口:本地 URL、preview、桌面应用、测试夹具或代码路径都可以;不要求生产已部署。
- 跑一次关键主链路,并按任务选择空态、错态、窄屏、刷新恢复、权限或异常输入等边界。
- 保存可复现步骤、截图/日志/控制台/测试输出。视觉任务才需要截图;纯代码审查不强制截图。
review-only 不授权业务副作用。支付、发消息、创建/删除真实记录、改权限等路径必须使用测试账号、夹具、mock 或 dry-run;没有安全替代时只做只读检查并列为未覆盖,不能借“审查”执行真实动作。
物证未就绪前不启动依赖它的审查。不要让 reviewer 猜一个稍后才可能出现的 evidence 文件。
2. 按任务选择审查视角
只选择有独立信息价值的 lane:
- 真实用户 / 产品:能否完成关键任务,路径成本、反馈、空错态、误触和恢复。
- 视觉 / 交互 / 可访问性:层级、可读性、响应式、键盘、对比、motion 和跨状态一致性。
- 实现 / 可靠性 / 安全:围绕本次 diff 与关键状态流找运行时错误、数据损坏、竞态、性能和维护风险。
运行时适配:
- Codex 有
collaboration.spawn_agent时,最多并行派出当前空闲槽允许的子 agent,主 agent保留裁决与修复职责。 - Claude Code 有
Agent能力时,用其原生并发方式。 - 没有可用 delegation 时,主 agent 顺序执行 lane;不得编造不存在的 agent 结果。
每条发现必须包含:
ID · 证据/复现 · 用户或系统影响 · severity · confidence · 最小修法 · 验证方法禁止只给分数、审美口号、名人类比或没有证据的“最佳实践”。
3. 主 agent 裁决
主 agent 独立复现并去重,不按投票机械采纳:
- 接受:证据可复现,影响真实,根因或修复边界清楚。
- 暂缓:问题真实,但超出授权、时间预算或会踩用户未提交改动。
- 拒绝:无法复现、重复、偏好冒充缺陷,或与用户/项目明确决策冲突。
优先级依次为:关键路径阻断或数据/安全风险 > 多视角指向同一机制 > 高置信局部修复 > 纯一致性和审美建议。
4. 按授权行动
- review-only:停在裁决,交付最多 5 条经过验证的发现。
- review-and-fix:只修已接受且在当前范围内的问题;不得顺手重构、版本 bump、deploy 或提交,除非用户或项目契约已授权。
- 修改后保持原项目的实现方式,避免用层层覆盖或无关重构绕过根因。
5. 复验
- 重跑原始复现步骤;视觉问题用相同 viewport、状态和数据做 before/after。
- 运行项目契约要求的静态、接口、端到端检查;有代码改动时覆盖至少 3 个相关边界 case。
- 检查工作区,确认未覆盖用户已有改动、未产生未授权提交或部署。
停止条件与输出
达到用户预算,或最高价值问题已完成裁决/修复并复验,就停止。不要为了填满角色配额继续制造低价值意见。
最终只交付:
- 审查范围与未覆盖范围;
- 接受、暂缓、拒绝的证据化结论;
- 若获授权,实际修改和复验命令输出;
- 尚需用户决策的高风险项。
不要输出 75–90 条 critique、综合评分或 token 消耗表,除非用户明确要求完整研究报告。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

