dbs-report

dbs-report is an AI skill for turning multiple saved business diagnosis states into one deliverable Markdown report. It solves the problem of valuable diagnosis conclusions being scattered across chat history by relying only on confirmed saved records created through dbs-save. Instead of summarizing from conversation memory, it reads stored Markdown snapshot files from the project’s session directory, merges them chronologically, removes duplicates, organizes the content, and produces a shareable report that can be archived or sent to partners, consultants, or stakeholders.

dbs-report supports generating reports for the current project, specifying a project with --slug, filtering saved records with --since YYYY-MM-DD, and combining both options. It validates whether enough saved records exist, sorts files by timestamp, parses frontmatter fields and six standard body sections, and writes a new Markdown report under the reports directory without overwriting previous outputs. The generated report includes complaint evolution, confirmed conclusions, rejected directions, unresolved issues, recommended next steps, and an appendix index of source files. It preserves traceability by tying report content back to saved files and avoids unsupported interpretation except for a brief evolution summary.

This skill is designed for users of dbskill who need a formal consulting-style artifact from accumulated diagnosis work. It is useful for business owners, consultants, founders, operators, or advisors who want to package confirmed diagnostic progress into a readable document for review, collaboration, or future reference. Its target use cases include preparing partner-facing summaries, reviewing multi-week diagnostic progress, comparing older and newer diagnosis versions, and creating a stable Markdown deliverable from previously saved diagnostic states.