Routing Optimization

基于调度日志数据收集、分析和A/B测试，动态优化routing规则以提升调度命中率和准确性。

Before installing or enabling this skill, verify the following: (1) Where and how dispatch_logs will be provided (files, DB, API) and ensure the agent is granted only the minimum read access necessary; (2) Confirm and review the implementations of helper functions referenced (match_route, get_base_priority, calculate_avg_accuracy, calculate_avg_time) so you know exactly how routing rules will be evaluated and modified; (3) Keep enable_auto_optimization disabled in production until you have validated results in staging, and require human approval for any automated rule changes; (4) Check that logs do not contain sensitive PII or credentials, and if they do, sanitize or restrict access; (5) Audit any integration code that actually writes routing changes to your systems — the SKILL.md is example code and does not show how writes are performed; (6) Because this is instruction-only, there are no bundled binaries or external downloads, but you should still monitor network activity from the environment where the skill runs to detect any unexpected exfiltration. If you need higher assurance, request concrete integration code or an implementation that clearly documents data flows and access controls.

Type: OpenClaw Skill Name: routing-optimization Version: 1.0.0 The routing-optimization skill bundle is designed to analyze and improve agent routing rules through hit-rate analysis, A/B testing, and accuracy metrics. The provided Python and PowerShell code in SKILL.md focuses on processing dispatch logs and generating optimization suggestions based on performance data. While the PowerShell function Optimize-RoutingRules contains a hardcoded path to a local configuration file, the logic is consistent with the stated purpose and lacks any indicators of malicious intent, data exfiltration, or unauthorized command execution.