Archestra 发布 OpenAPPA 安全引擎，两大基准攻击成功率为 0%
原标题：New Archestra's OpenAPPA Saturates Two Major Security Benchmarks with a 0% Attack Success Rate
AI 摘要
Archestra 发布开源安全引擎 OpenAPPA，用于阻止提示注入或模型幻觉导致的数据外泄。该引擎运行在智能体提示与执行循环之外，通过 appa.toml 配置数据源、受众、信任等级和权限，并采用 Agentic Permissions Policy Algebra（APPA）实现确定性策略执行。在 Bench-Corp 和 AgentThreatBench 基准上，OpenAPPA 报告 0% 攻击成功率、89% 任务完成率，优于 Claude Code 自动模式（10% 攻击成功率、90% 完成率）和 Microsoft FIDES（31% 攻击成功率、41% 完成率）。目前 OpenAPPA 处于预览阶段，形式化代数与恢复保证已发布在 arXiv。
正文节选
Archestra released OpenAPPA, an open-source security engine designed to stop data exfiltration caused by prompt injection or model hallucination. OpenAPPA runs outside the agent’s prompt and execution loop. Its configuration details concepts such as data sources, audiences, trust levels, and authorities, along with their associated deterministic security enforcement rules. The team reports zero successful attacks when running security benchmarks Bench-Corp (20 multi-step enterprise workflows) an