Outage Resilience Pro — v1.0 Release Notes
Released: 2026-04-14
Price: $4.99
Category: orchestration
What It Does
Outage Resilience Pro is a production resilience suite that keeps your OpenClaw agent pipelines running when LLM providers go down, hit rate limits, or return errors. It monitors provider health in real-time, automatically fails over to backup models when your primary provider stumbles, and preserves pipeline state so you can recover exactly where you left off — no lost work, no silent failures.
Think of it as a circuit breaker for your agent: when Anthropic or OpenAI has an outage, your workflows keep running on a backup provider instead of crashing.
Use Cases
- Production agent operators who run cron-driven pipelines and can't afford silent failures when a provider hits a 429 or goes down at 3am
- Teams using multiple LLM providers (OpenRouter, Anthropic, OpenAI, Google) who want automatic failover instead of manually swapping API keys during incidents
- Anyone who has lost pipeline progress during an outage and wants timestamped state snapshots they can restore from, rather than restarting from scratch
Requirements
- OpenClaw — any recent version
- Python 3 — standard library only (urllib, json, os, pathlib)
- No additional API keys required — uses your existing provider credentials already in OpenClaw
- No external dependencies — queue manager, state snapshots, and failover config are all local filesystem
Example Usage
Check provider health:
Check LLM provider health status
✅ OpenRouter 🟢 healthy (123ms)
⚠️ Anthropic 🟡 unknown (no key configured)
⚠️ OpenAI 🟡 unknown (no key configured)
⚠️ Google 🟡 unknown (no key configured)
1/1 configured providers healthy
Set up failover chains:
Make my agent pipeline resilient to outages
Failover configuration saved → resilience-config.json
Chain 1: openrouter → anthropic → openai
Chain 2: anthropic → openrouter → google
Triggers: 3 consecutive failures OR latency > 10s OR 5x 429s
Cooldown: 5 minutes before retrying failed provider
Budget guard: max 2x normal cost during failover
Save and restore pipeline state:
scripts/resilience_manager.py state --save
Snapshot created: resilience-snapshots/snapshot-2026-04-14T17-28-00/
Files captured: 53 (memory/*.md, workspace files)
Manifest: manifest.json
scripts/resilience_manager.py state --list
snapshot-2026-04-14T17-28-00 (53 files)
snapshot-2026-04-13T09-15-42 (51 files)
scripts/resilience_manager.py state --restore snapshot-2026-04-14T17-28-00
Pipeline state restored ✅