Willoscar

This skill creates papers/extraction_table.csv from included papers following the protocol schema and preserves provenance in auditable form.

This skill de-slots and harmonizes paper voice across sections markdowns while preserving meaning and citations.

This skill initializes a new workspace by copying the standard artifact template into the target directory and ensuring core files exist.

This skill expands and refines search keywords to improve queries.md recall and precision while keeping scope under control.

This skill writes evidence-backed survey drafts from approved outlines and citations, producing coherent, thesis-driven prose that integrates multiple sources.

This skill injects budget-compliant citations into drafts without adding new facts, ensuring global uniqueness while preserving scope.

This skill enforces argument continuity by generating a ledger and self-checks to ensure premise hygiene before merge.

This skill writes a structured synthesis of evidence from papers/extraction_table.csv, detailing themes, biases, and limitations for transparent systematic

This skill polishes one sections/S subfile into survey-grade prose by refining argument flow while preserving citations for reliable pre-merge polishing.

This skill generates non-prose visuals for surveys, creating timeline milestones and figure specs grounded in citations for easy verification.

This skill converts a taxonomy into a bullets-only, checkable outline with sections and subsections, ensuring each subsection has at least three actionable

This skill rewrites the opening paragraph of flagged sections to replace overview narration with content-bearing framing while preserving citations.

This skill generates an auditable screening_log.csv from the protocol and candidate lists, producing transparent inclusion decisions with protocol-grounded

This skill builds a prerequisite concept graph from a tutorial spec and saves it as outline/concept_graph.yml to guide sequencing.

This skill bootstrap taxonomy seeds from existing survey papers to accelerate taxonomy-builder alignment and scoping.

This skill builds a novelty matrix aligning each claim against closest prior work, detailing overlaps, deltas, and evidence.

This skill provides an auditable planner pass for outlines by reporting coverage, reuse hotspots, and axis specificity without prose.

This skill crafts concise H2 lead blocks that unify chapter subsections by presenting the lens, contrasts, and calibration without adding facts.

This skill deterministically merges per-section prose into a cohesive DRAFT.md, preserving outline order and transitional text from outline/transitions.

This skill converts a manuscript into auditable claims with precise source pointers, separating empirical and conceptual statements for traceable review.

This skill writes cohesive TUTORIAL.md modules from approved plans, preserving running example continuity, exercises with verification, and answer outlines.

This skill analyzes evidence artifacts and outputs an actionable upstream TODO plan to fix gaps and unblock writing.

This skill produces concise, paper-like transitions that connect adjacent subsections without adding facts, preserving coherence from outlines to drafts.

This skill rewrites outline.yml to a paper-like budget, consolidating H2s and H3s to 6–8 major sections while preserving scope.

This skill converts each H3 subsection into a writable intent card with concrete, non-copyable briefs guiding evidence-based drafting.

This skill turns a concept graph into a teachable module plan with objectives, outputs, and running steps saved as outline/module_plan.yml.

This skill audits each claim's evidence by inspecting the manuscript and reporting gaps with actionable fixes in MISSING_EVIDENCE.md.

This skill downloads PDFs when available and extracts full text to enrich evidence for papers, storing results in papers/fulltext.

This skill rewrites the claim-evidence matrix as a data-driven index using evidence packs, ensuring no prose and citation-backed claims.

This skill writes a complete referee report grounded in extracted claims and evidence gaps, identifying actionable fixes for novelty, soundness, clarity, and

This skill crafts a survey-quality paragraph from evidence packs, balancing tension, contrast, evaluation anchor, and limitation for subsection prose.

This skill analyzes post-merge draft content to detect generator voice from outline transitions and routes fixes to the earliest source.

This skill writes per-section survey prose into sections/*.md files by enforcing citation scope and argument moves for QA-ready drafts.

This skill builds a large, verifiable candidate pool of literature with stable IDs and provenance for evidence-first surveys.

This skill generates an executable systematic review protocol with databases, queries, inclusion/exclusion, and audit-ready screening steps.

This skill binds section-specific evidence IDs from the evidence bank to each subsection, enabling precise writer-auditor references.

This skill downloads arXiv agent-survey PDFs, extracts text, and generates a STYLE_REPORT to help you study real survey structures for better writing.

This skill scaffolds a minimal LaTeX project from an approved Markdown draft and bib file, producing latex/main.tex.

This skill runs end-to-end research pipelines with workspaces and checkpoints, delivering auditable Units.csv and artifacts.

This skill fills outline/tables_index.md from evidence packs with citation-backed rows to support later writing and traceability.

This skill audits citation anchoring to ensure citation markers stay within the same ### subsection after polishing, preventing drift.

This skill generates structured, citation-ready notes for each core paper, capturing method, results, and limitations for evidence-based writing.

This skill designs both index and appendix table schemas with evidence-backed columns, ensuring fillability and verifiability before publication.

This skill helps you build a diverse, explainable paper-to-subsection mapping for outline coverage by producing outline/mapping.tsv and diagnostics.

This skill writes paper-like front matter sections (Abstract, Introduction, Related Work, Discussion, Conclusion) with high citation density for a cohesive

This skill delivers a one-page literature snapshot in output/SNAPSHOT.md via a bullet-first outline and core papers, enabling fast, actionable reading.

This skill ensures numeric claims are reviewer-safe by enforcing minimal protocol context (task, metric, constraint) and avoiding underspecified model naming.

This skill defines a tutorial spec with audience, prerequisites, measurable objectives, non-goals, and a running example for deterministic planning.