What problem does it solve? Literature-ingest skills leave research-wiki paper pages as bare scaffolds with 10 unfilled TODO sections, so the knowledge base sits empty until someone manually reads every paper. This Skill automates that back half: it reads each scaffolded page, fetches paper content from external sources, and writes concise prose summaries into every TODO section. ## Core Features & Use Cases - Batch TODO filling: Targets a single paper by slug, only papers with remaining TODOs (default), or all papers with --force, capped at 20 per run to control token budget. - 5-source fallback fetch chain: Tries alphaxiv overview, alphaxiv abstract, deepxiv brief, arXiv API abstract, then the page's own stored abstract, recording provenance for every paper. - Protected-section safety: Never touches YAML frontmatter, the graph-generated ## Connections section, or the immutable ## Abstract (original) block; idempotent re-runs only fill still-TODO sections. - Use Case: After /research-lit "kv-cache optimization" ingests 15 papers as scaffolds, run /wiki-enrich to fill all 15 pages with thesis, method, results, limitations, and reusable-ingredient summaries, then feed them into /idea-creator for brainstorming. ## Quick Start Run /wiki-enrich to fill every research-wiki paper page that still has TODO sections, or /wiki-enrich vllm to enrich a single paper by slug.
