Overview
Production agent pipelines fail — providers go down, rate limits hit at 3am, and a single 429 can stall your entire workflow for hours. Outage Resilience Pro wraps your OpenClaw pipelines with automatic detection, intelligent failover, and state preservation so your agents keep running even when your primary LLM provider doesn't.
The skill monitors provider health in real time, queues requests during outages instead of dropping them, and restores pipeline state cleanly once a provider recovers. Built for teams running cron-driven agent workflows where downtime has a real cost.
Use Cases
Auto-Failover SetupConfigure primary and backup model chains that switch automatically on 429s or 5xx errors.
Provider Health MonitoringReal-time checks on OpenAI, Anthropic, Gemini and OpenRouter — know before your crons do.
Rate Limit HandlingQueue requests intelligently during rate limit windows instead of crashing the pipeline.
Pipeline State RecoveryPreserve in-progress agent state so pipelines resume from where they stalled, not from scratch.
Outage AlertingGet notified the moment a provider degrades — before it becomes a missed cron run.
Fault-Tolerant CronsWrap any scheduled agent job with resilience so it always completes, even on a bad provider day.
Requirements
- OpenClaw v2026.x or later
- Python 3.9+
- At least one configured LLM provider via OpenRouter or direct API key
- Optional: secondary provider API key for failover chain
Example Usage
Command
Check LLM provider health status
Make my agent pipeline resilient to outages
Set up failover from claude-sonnet to gemini-pro with request queuing