Infra Operator

Watch production, find the fault, operate the edge — under guardrails.

For: Teams running an agent across their edge, database, and error telemetry.

Sentry Telemetry
Errors, traces, and release health — the first read on what's wrong.
unknown
Cloudflare Edge & network
Workers, DNS, R2, and observability — operate the edge.
healthy
Neon Database
Serverless Postgres branches, migrations, and queries.
healthy
GitHub Change control
Land fixes as reviewed PRs, never ad-hoc edits.
healthy

This kit touches live infrastructure, so its writes are the highest-stakes on the site — scope them tightly and require approval. Reads (metrics, traces, schemas) are the safe default the agent should live in; every change goes through review, not ad-hoc edits.