跳转至

Eval 命令

管理由评估驱动的开发流程。

用法

/eval [define|check|report|list] [feature-name]

定义评估

/eval define feature-name

创建新的评估定义:

  1. 使用模板创建 .claude/evals/feature-name.md:
## EVAL: feature-name
Created: $(date)

### Evals de Capacidade
- [ ] [Descrição da capacidade 1]
- [ ] [Descrição da capacidade 2]

### Evals de Regressão
- [ ] [Comportamento existente 1 ainda funciona]
- [ ] [Comportamento existente 2 ainda funciona]

### Critérios de Sucesso
- pass@3 > 90% para evals de capacidade
- pass^3 = 100% para evals de regressão
  1. 请用户填写具体标准

检查评估

/eval check feature-name

为某个功能运行评估:

  1. 读取 .claude/evals/feature-name.md 中的评估定义
  2. 对于每个能力评估:
  3. 尝试验证标准
  4. 记录 PASS/FAIL
  5. 将尝试保存到 .claude/evals/feature-name.log
  6. 对于每个回归评估:
  7. 运行相关测试
  8. 与基线进行比较
  9. 记录 PASS/FAIL
  10. 报告当前状态:
EVAL CHECK: feature-name
========================
Capability: X/Y passing
Regression: X/Y passing
Status: IN PROGRESS / READY

评估报告

/eval report feature-name

生成完整的评估报告:

EVAL REPORT: feature-name
=========================
Generated: $(date)

CAPABILITY EVALS
----------------
[eval-1]: PASS (pass@1)
[eval-2]: PASS (pass@2) - required retry
[eval-3]: FAIL - see notes

REGRESSION EVALS
----------------
[test-1]: PASS
[test-2]: PASS
[test-3]: PASS

METRICS
-------
Capability pass@1: 67%
Capability pass@3: 100%
Regression pass^3: 100%

NOTES
-----
[Any issues, edge cases, or observations]

RECOMMENDATION
--------------
[SHIP / NEEDS WORK / BLOCKED]

列出评估

/eval list

显示所有评估定义:

EVAL DEFINITIONS
================
feature-auth      [3/5 passing] IN PROGRESS
feature-search    [5/5 passing] READY
feature-export    [0/4 passing] NOT STARTED

参数

$ARGUMENTS: - define <name> - 创建新的评估定义 - check <name> - 运行并检查评估 - report <name> - 生成完整报告 - list - 显示所有评估 - clean - 删除旧的评估日志(保留最近 10 次运行)

本页原文 Markdown:在 AtomGit 查看·内容源自开源项目 ev/ECC