Report to PDF

Most documents fail at the executive summary because the writer drafts it first, before they know what the rest actually says. This skill writes it last, on a formula — sentence one is context, sentence two is the single most important finding, sentences three and four are the supporting points, sentence five is the recommendation — so the summary stands alone. A reader who only sees that paragraph still understands the situation, the findings, and what should happen next. That matters because for most documents, the executive summary is the document.

Structure scales to length. A 1-2 page brief uses title, executive summary, three to five key findings, two to three recommendations, next steps. A 3-5 page version adds background, methodology if research-based, findings organised by theme rather than chronology, analysis, and a recommendations table. A 10+ page comprehensive version adds a table of contents, expanded methodology, multi-section findings with sub-headers, prioritised recommendations, implementation plan, and appendices for supporting data. The skill asks how long you want it and matches detail to that — and if the input is sparse, it produces a shorter honest write-up rather than padding to a target length.

Findings and recommendations are kept in separate sections. Mixing "what we found" with "what we think you should do" makes both harder to read and lets weak recommendations hide behind strong data. Every recommendation in the output names an owner and a timeline — "Product team should redesign onboarding by Q3" instead of "we recommend improving the onboarding process." Every number gets context: 23% of what, compared to what, over what period. The "so what" test runs after every finding — if a finding has no implication for action, it either gets rewritten or paired with one.

The audience calibration matters. Board and investors get formal, metric-heavy, risk-aware, concise prose with no jargon. Executives get strategic, decision-focused, one-page-preferred output. Technical teams get longer, data-rich, methodology-included write-ups. External clients get professional, jargon-free, results-focused content with every recommendation explained. The skill asks who's reading before it writes. Output is clean Markdown ready to export to PDF, with headers, tables, and white space designed for scanning first and reading second — because most readers scan, and the document needs to tell the story on its skeleton before they commit to the paragraphs.