wuyoscar/ISC-Bench

• highreadme#1

  Reposition README's opening to highlight 'sensitive data generation'

  原因：

  当前

  <h2 align="center">Internal Safety Collapse in Frontier Large Language Models</h2>

  复制粘贴的修复

  复制

  <h2 align="center">Internal Safety Collapse: Turning LLMs into Sensitive Data Generators</h2>
  <p align="center">A benchmark for evaluating LLM and AI agent robustness against internal safety collapse and sensitive information leakage.</p>

  待办完成忽略
• mediumtopics#2

  Add more specific topics related to data leakage and privacy

  原因：

  当前

  agent-safety, ai-safety, benchmark, jailbreak, large-language-models, llm-safety, red-teaming, safety-evaluation

  复制粘贴的修复

  复制

  agent-safety, ai-safety, benchmark, data-leakage, data-privacy, jailbreak, large-language-models, llm-safety, privacy-breach, red-teaming, safety-evaluation, sensitive-data

  待办完成忽略
• mediumlicense#3

  Clarify the existing license in the README

  原因：

  复制粘贴的修复

  复制

  ## License
  This project is released under [describe your specific license(s) here, e.g., "a custom license combining elements of X and Y"]. Please refer to the [LICENSE file](LICENSE) for full details.

  待办完成忽略

本次扫描解析到的品类 GEO 通道：google/gemini-2.5-flash, deepseek/deepseek-v4-flash

• 品类问题

  How to benchmark large language models for potential safety vulnerabilities and data generation risks?

  你：未被推荐
  AI 推荐顺序：

  1. Giskard
  2. MLCommons MLPerf Inference
  3. OpenAI Evals
  4. Anthropic's Constitutional AI
  5. Hugging Face Evaluate Library
  6. Perspective API
  7. AdvBench
  8. Gauntlet

  AI 推荐了 8 个替代方案，却始终没点名 wuyoscar/ISC-Bench。这就是要补上的差距。

  查看 AI 完整回答
• 品类问题

  Seeking tools to test AI agent robustness against internal safety collapse and sensitive information leakage.

  你：未被推荐
  AI 推荐顺序：

  1. Giskard (GiskardAI/giskard)
  2. Robust Intelligence (RI Platform)
  3. Adversarial Robustness Toolbox (ART) (Trusted-AI/adversarial-robustness-toolbox)
  4. DynoSafe
  5. OWASP Top 10 for LLMs
  6. Fiddler AI

  AI 推荐了 6 个替代方案，却始终没点名 wuyoscar/ISC-Bench。这就是要补上的差距。

  查看 AI 完整回答

• Metadata completeness

  pass

• README presence

  pass

• Compared to common alternatives in this category, what is the core differentiator of wuyoscar/ISC-Bench?

  pass
  ✓AI 明确点名了 wuyoscar/ISC-Bench

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• If a team adopts wuyoscar/ISC-Bench in production, what risks or prerequisites should they evaluate first?

  pass
  ✓AI 明确点名了 wuyoscar/ISC-Bench

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• In one sentence, what problem does the repo wuyoscar/ISC-Bench solve, and who is the primary audience?

  pass
  ✓AI 明确点名了 wuyoscar/ISC-Bench

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？