diagnose-openclaw-model-routing-and-fallback-usage用于分析 OpenClaw 中“某个 provider/模型为什么看起来没被用上”、主模型与 fallback 实际是否生效、以及 ClaudeCodeCLI 协作任务为何中断。遇到这些情况就应触发：用户提到“没请求打到 local-router”“明明注册了模型却没流量”“Agent 到底在用哪个模型”“fal...
Install via ClawdBot CLI:
clawdbot install can4hou6joeng4/diagnose-openclaw-model-routing-and-fallback-usageGrade Fair — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Sends data to undocumented external endpoint (potential exfiltration)
Post → https://chatgpt.com/backend-api/codex/responsesAccesses system directories or attempts privilege escalation
/var/log/Calls external URL not in known-safe list
http://localhost:8317/v1`AI Analysis
The skill's primary function is diagnostic analysis of local routing and fallback configurations, which aligns with its stated purpose. The external endpoint 'http://localhost:8317/v1' is a local service consistent with the skill's routing analysis, not an unauthorized external server. The other flagged signals appear to be incidental references within diagnostic examples rather than active malicious behavior.
Generated May 20, 2026
当用户怀疑某个模型未被实际调用时，通过分析配置和日志，区分 provider 级别和模型级别的流量，定位问题根源。
用户需要确认主模型失败后 fallback 是否生效，通过检查 fallback 链配置和日志中的 fallback 决策记录，给出严谨结论。
当 ClaudeCodeCLI 协作任务异常退出（如 code 143），通过退出码解读、执行日志分析，判断是外部终止还是内部错误，并提供规避方案。
用户感觉回答风格不符预期，通过检查主 Agent 及辅助组件的模型绑定配置，确定实际使用的模型链，消除认知偏差。
当旧日志显示某模型曾被调用，但当前未使用时，帮助区分历史轨迹与现状，防止误判当前链路。
提供针对 AI Agent 模型路由、fallback 行为及 CLI 协作中断的专业诊断服务，帮助用户快速定位配置或运行时问题。
整合配置与日志，提供可视化的模型路由和 fallback 使用分析，支持自定义报告，帮助优化模型选择与成本。
针对频繁的 CLI 中断问题，提供执行流程重构、超时策略设计、健康探针部署等咨询服务，提升任务稳定性。
💬 Integration Tip
将诊断步骤自动化到运维脚本中，定期执行配置审计与日志分析，并在出现典型错误模式时主动告警。
Scored Jun 29, 2026
Audited Apr 17, 2026 · audit v1.0
Use the ClawdHub CLI to search, install, update, and publish agent skills from clawdhub.com. Use when you need to fetch new skills on the fly, sync installed skills to latest or a specific version, or publish new/updated skill folders with the npm-installed clawdhub CLI.
Mission control dashboard for OpenClaw - real-time session monitoring, LLM usage tracking, cost intelligence, and system vitals. View all your AI agents in o...
Transcribe YouTube videos to text by extracting captions and subtitles directly from the video URL using yt-dlp without audio processing.
Manage a self-hosted Trello-like board via `wekancli`. Create, move and archive cards, lists and boards on a WeKan server. Use when user asks about task boar...
Proactive security monitoring, threat scanning, and auto-remediation for OpenClaw deployments
Create or improve SOUL.md files for OpenClaw agents through guided conversation. Use when designing agent personality, crafting a soul, or saying "help me create a soul". Supports self-improvement.