`Superlog`, self-installing observability with bug-fix PRs

Observability setup moves from dashboard wiring to repo-scanning instrumentation. The promise is strong for small teams, but pricing and real-world accuracy decide whether it replaces Sentry or Datadog.

• —The wizard scans a repo and adds structured logs, traces, and metrics through OpenTelemetry; setup pain is the main wedge.
• —Incidents are fingerprinted and grouped, so alert flow becomes one issue instead of duplicate spam in Slack.
• —The agent investigates errors and opens a tested PR when context is enough; otherwise it posts findings and pulls in likely owners.
• —Cost tracking includes LLM/upstream spend by callsite, tenant, and model, useful for multi-tenant SaaS margins.