The writer-context-pack skill acts as a deterministic bridge between raw research data and final manuscript drafting. By merging subsection briefs, evidence drafts, and anchor facts into a single per-subsection JSONL, it eliminates 'hollow' AI writing and ensures every paragraph is grounded in verified citations and specific evaluation metrics. It enforces a strict 'writer contract' that mandates comparison-based writing and prevents repetitive template-like narration, resulting in professional, publication-ready prose that adheres to specific evidence-first constraints.
主要功能
01Enforces 'Anti-Template' rules to prevent repetitive AI narration and meta-signposting
02Restricts citation scope to pre-mapped bibliography keys to prevent reference hallucinations
03Integrates Paper-Voice palettes to ensure varied, professional academic tone and style harmonization
04Consolidates research briefs, evidence drafts, and anchor sheets into a single deterministic source
0596 GitHub stars
06Mandates the use of specific comparison cards and numeric evaluation anchors for grounded claims
使用场景
01Ensuring large-scale technical documentation maintains a consistent evidence density across diverse contributors
02Automating the transition from raw research notes to structured, argumentative academic prose
03Bridging the gap between literature review artifacts and the first draft of a technical paper