Automation & AI Workflow SRE specializing in production reliability, incident post-mortems, and silent-failure prevention across n8n, Dify, Activepieces, Make, Shopify Webhooks, and Docker.
grinem/hooktriage(MIT) — Zero-Trust CLI & MCP Server (npx -y github:grinem/hooktriage scan/--mcp, 0 bytes uploaded).grinem/n8n-ai-watchdog(MIT) — Zero-Trust 3-node n8n Error Workflow with local secret scrubbing (API keys, JWTs, Bearer tokens, PII) and root-cause alert classification for429,401/403, webhook drops, and empty AI outputs.grinem/n8n-production-incident-fixes— Live engineering archive of 100+ diagnosed production failures & verified recovery patterns.- 🌐 Live Knowledge Base & Client-Side Audit: https://hooktriage.com (in-browser workflow static audit + case study archive)
- Silent AI Agent & LLM Drops: Workflows ending "green" while returning
[], truncated JSON, or hitting token/tool loops. - Shopify / Stripe Webhook Silences: 5-second webhook cutoffs, missing idempotency keys, and HMAC verification drift.
- OAuth2 & Rate-Limit Deadlocks: Concurrent n8n/Make token refresh collisions and HTTP
429cascading retries. - Self-Hosted Worker Reliability: Postgres/Redis queue exhaustion, reverse-proxy timeout disconnects, and memory leaks.