source-post

# Source Post Use this skill when source material must become a publishable source post, not just a chat summary. ## Workflow 1. Capture the source before writing: title, author, date, canonical URL, access date, article structure, tables, footnotes, glossary, links, images, charts, video, downloads, and known access limits. 2. Set the fidelity boundary. Preserve claims, structure, numbers, terms, links, and media references. Do not publish a full copyrighted repost or full translation unless the user owns the rights or provides licensed source text. 3. Define the reader job before drafting: what should a human understand in the first minute, what structure helps them, and what decision or project implication they should leave with. 4. Write the reader-facing blog before the evidence apparatus. Lead with a short article narrative, then the simplest source structure needed for orientation, then analysis. Put source ledger, copyright boundary, and validation details after the reader understands the source. 5. Use `effective-interact` only as a secondary summary or visualization layer. Do not use a raw generated interaction report as the blog body. 6. Add a project-iteration review: what the source confirms, what should change in this repo, what was already changed, what may be redundant, what should stay out of scope, and what needs a later SDD/spec. 7. Validate the generated post with `scripts/validate-source-post.mjs`, the normal source-post validation, a rendered screensh...