AI Retainer Health Check is a skill for AI automation agencies that transforms unstructured monthly inputs — goals, completed work, blockers, and risks — into a standardized client-facing Retainer Health Check. It calculates a health score, categorizes blockers by type (such as technical or access-related), prioritizes client actions, and generates a strategic plan for the upcoming month.
The skill produces five output formats simultaneously: PDF, DOCX, Markdown, CSV, and JSON. This means the same report run produces a client-ready document, a machine-readable data export for tracking trends over time, and a version suitable for embedding in other workflows. The JSON and CSV exports enable programmatic tracking of retainer health across multiple clients over successive months.
A stated design constraint is that the skill operates deterministically and avoids generating invented ROI figures or fabricated metrics — outputs are derived strictly from the inputs provided. This is relevant for agencies that need defensible, evidence-based reporting rather than AI-generated estimates dressed as data.
The skill is intended to replace manual report formatting workflows, removing the need for account managers or developers to maintain and populate Word templates by hand. It can be triggered via agent prompt or CLI, which makes it composable with broader automation pipelines. It is not a data analytics platform and does not pull live data from external systems; inputs must be supplied by the caller.