survey-builder
Survey Builder
Use this to turn a paper set into a structured survey.
For a compact Related Work section inside an ongoing project, use ai-research-workflow/references/project-workflow.md and its related-work matrix. Link existing paper cards and record project relevance, evidence, code and reproduction status. Apply this skill's full survey/tree deliverables when the user requests a literature survey/research map; do not expand a project index into that larger deliverable automatically.
Canonical Paper Card Gate
When a survey includes paper cards, also use paper-card-delivery. That skill is the canonical standard for official-source verification, fixed card format, figure/caption selection, sorting, and validation. Survey Builder must not finalize cards from its local summary rules alone.
Canonical Chinese Technical Writing Gate
For Chinese survey prose, taxonomy descriptions, challenge-insight explanations, paper matrices, reading routes, figure captions, and non-bilingual mind-map/table labels, also use chinese-technical-writing. Keep required bilingual mind-map nodes as English newline Chinese when specified, but write ordinary Chinese explanation in Chinese-first terminology.
Canonical Platform Delivery Gate
Use research-doc-workflow and the matching platform adapter (feishu-doc-workflow, notion-doc-workflow, or obsidian-doc-workflow) to deliver the same Markdown-first survey structure to Feishu/Lark, Notion, or Obsidian. Keep headings, tables, lists, links, and prose identical; adapt only tree widgets, page/database hierarchy, captions, properties/frontmatter, and verification. Do not require a Feishu page when the user supplied a Notion or Obsidian target.