jnmetacode avatar

Verification Before Completion

This is a discipline enforcer that stops you from saying "tests pass" or "bug fixed" without running the actual verification command in the same session. The core rule is simple: no fresh evidence, no completion claim. It includes a table of what counts as real verification versus what doesn't (spoiler: linter passing doesn't prove your build works), plus a list of red flags like using "should work" or trusting an agent's success report. Born from 24 documented failures where unverified claims broke trust and shipped broken code. Use it before any commit, PR, or completion statement. It's ruthlessly specific about the difference between confidence and proof.

skills.sh installs
1.1K
GitHub stars
8.5K
last updated
Aug 21, 2026

Install

$ npx -y skills add jnmetacode/superpowers-zh --skill verification-before-completion --agent claude-code

Installs into .claude/skills of the current project.

Files

完成前验证

概述

核心原则: 始终用证据支撑结论。

对这条规则敷衍了事,就等于违背了它的精神。

铁律

没有新鲜的验证证据,不许宣称完成

如果你在这条消息中没有运行验证命令,就不能声称测试通过。

门控函数

在宣称任何状态或表达满意之前:

1. 确定:什么命令能证明这个结论?
2. 运行:执行完整命令(重新运行,完整执行)
3. 阅读:完整输出,检查退出码,统计失败数
4. 验证:输出是否支持这个结论?
   - 如果否:用证据说明实际状态
   - 如果是:带证据陈述结论
5. 只有这时:才能做出结论

跳过任何一步 = 说谎,不是验证

常见失败模式

结论需要不够格
测试通过测试命令输出:0 failures之前的运行结果、"应该会通过"
Linter 无报错Linter 输出:0 errors部分检查、推断
构建成功构建命令:exit 0linter 通过、日志看起来没问题
Bug 已修复测试原始症状:通过代码改了,假设已修复
回归测试有效红-绿循环已验证测试只通过了一次
代理已完成VCS diff 显示变更代理报告"成功"
需求已满足逐项核对清单测试通过

红线——停下来

  • 使用"应该"、"大概"、"似乎"
  • 验证前就表达满意("太好了!"、"完美!"、"搞定!"等)
  • 即将提交/推送/创建 PR 却没有验证
  • 信任代理的成功报告
  • 依赖部分验证
  • 想着"就这一次"
  • 累了想赶紧收工
  • 任何暗示成功但实际未运行验证的措辞

防止合理化

借口现实
"应该能行了"运行验证命令
"我有信心"信心 ≠ 证据
"就这一次"没有例外
"Linter 通过了"Linter ≠ 编译器
"代理说成功了"独立验证
"我累了"疲劳 ≠ 借口
"部分检查就够了"部分检查什么也证明不了
"换个说法这条规则就不适用了"精神大于字面

关键模式

测试:

✅ [运行测试命令] [看到:34/34 pass] "全部测试通过"
❌ "应该能通过了" / "看起来对了"

回归测试(TDD 红-绿):

✅ 编写 → 运行(通过)→ 回退修复 → 运行(必须失败)→ 恢复 → 运行(通过)
❌ "我写了回归测试"(没有经过红-绿验证)

Sponsors

Details

Repository
github.com/jnmetacode/superpowers-zh
First seen
Jun 3, 2026

More from jnmetacode/superpowers-zh

All 20

Recommended

1juliusbrussee avatarcavemanjuliusbrussee/cavemanUltra-compressed communication mode. Cuts output tokens 65% (measured) by speaking like caveman while keeping full technical accuracy. Supports intensity levels618.3Kskills.sh installs
2mattpocock avatargrill-memattpocock/skillsA relentless interview to sharpen a plan or design.1.5Mskills.sh installs
3shadcn avatarimproveshadcn/improveSurvey any codebase as a senior advisor and produce prioritized, self-contained implementation plans for OTHER models/agents to execute. Strictly read-only on s40Kskills.sh installs
4obra avatarsystematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes326.7Kskills.sh installs
5forrestchang avatarkarpathy-guidelinesforrestchang/andrej-karpathy-skillsBehavioral guidelines to reduce common LLM coding mistakes through explicit assumptions, simplicity, and verifiable success criteria.17.9Kskills.sh installs
6vercel-labs avatarfind-skillsvercel-labs/skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express int4.3Mskills.sh installs