paper-memory-builder

To create persistent memory from a manuscript, this Skill reads your drafts and figures to generate structured YAML files (.paper/claims.yml and .paper/figures.yml). This memory layer lets AI reuse context during writing and auditing without rereading the manuscript.

A manuscript memory layer works by translating drafts, figures, and Zotero metadata into structured YAML files. It preserves the original manuscript unchanged while delegating context storage to .paper/claims.yml and .paper/figures.yml for revision workflows.

Yes, you can extract claims without losing data because the Skill preserves existing memory and enforces anti-leakage rules. It writes structured data to a dedicated memory layer while keeping your original manuscript drafts and figure assets completely unchanged.

The best way to prepare a paper is to build a persistent memory layer by running this Skill on your manuscript folder. It generates .paper/claims.yml and .paper/figures.yml from your drafts and figures to support continuous writing and audit passes.

You should use a dedicated memory layer to avoid the overhead of having AI reread the manuscript each time. By translating drafts and figures into structured YAML files, AI-assisted writing, revision, and auditing workflows can efficiently reuse the context.

You should not use a manuscript memory builder to summarize cited papers in a literature cluster. This Skill specifically translates your own manuscript drafts and figure assets into structured YAML memory for AI-assisted writing and auditing.