使用方法
- 复制安装命令，在自己的智能体环境安装前先检查源码包。
- 从原始文档核对所需工具、凭据和支持的宿主。
- 说明输入材料与期望交付物，并对照来源检查输出。
npx skills add https://github.com/yogsoth-ai/de-anthropocentric-research-engine --skill validate-mitigation-effect开始前需了解
- 安装不包含模型订阅、API 额度或访问文件的权限。
- 仓库星数和安装量反映活跃情况，不等于独立质量或安全测试。
仓库原始内容
按源仓库内容呈现，保留标题、案例、代码、表格、链接以及原文引用的演示图片。
validate-mitigation-effect
Purpose
Re-run the relevant failure or adversarial test after mitigation and verify actual risk reduction without unacceptable new failures.
Input contract
yaml
required: [failure_mode, baseline_risk, mitigation, validation_test]
optional: [new_failure_conditions, residual_risk_scale]
constraints: [baseline and post-mitigation tests must be comparable]Procedure
- Reproduce the baseline failure or risk measurement.
- Apply the mitigation and re-run the relevant test.
- Compare residual risk and inspect new failure modes.
Output contract
yaml
produces: [validation_result, residual_risk, new_failure_modes, mitigation_decision]
delta_fields: [findings, evidence_updates, uncertainties, decisions]Quality gates
- Mitigation passes only when risk decreases under comparable testing and no unacceptable new mode appears.
Failure and counterexamples
Do not infer mitigation effectiveness from implementation alone.
Provenance map
- resolved: mitigation-validation
