Add dashboard for pipeline infrastructure reliability

We need to be able to detect downtime in pipeline infrastructure:
- Vault access
- Network reliability
- Central service reliability (Azure artifacts / nuget / maven central...)

We should be able to detect a decrease in success rate across "most stable" pipelines. i.e. Of the pipelines that failed the least in the last X days, the success rate decreased by more than Y in the last day.

Start by defining the dashboard scope from the requested Vault, network, and central service reliability signals. Clarify the pipeline population, time window X, threshold Y, data sources, and the expected dashboard before locating an entry point; done means downtime and declining success rates are detected for the specified services and stable pipelines.

Written by the indexing model from the issue text.