Claude Skills by WissingChen

Top-tier CS/AI/Robotics manuscript writing taste. Load when drafting, revising, or polishing academic papers in computer vision, machine learning, robotics, image processing, and related fields. Not for research reports or slides — those belong to write_md and slide_deck.

Direction-level experiment execution, expectation tracking, and artifact workflow. Load when a user starts, runs, checkpoints, or closes experiments. Requires each run to target the direction's current central question, enforces convergence accounting, assembles reports with verified visuals, and owns branch/worktree lifecycle, output hygiene, and promotion. Evidence interpretation belongs to result_analysis; visual production to result_visualization.

Knowledge-base retrieval, citation-graph expansion, and capture for papers and external sources. Load when the user asks to search literature, look up a paper, summarize a paper into the knowledge base, or reuse previously found references. Enforces local-first retrieval, forward citation chasing from key papers, a query log, and mandatory capture. Judging whether a literature gap is real or a direction is worth doing belongs to research_progress.

Plain-talk output enforcer. MUST be loaded at the start of every session, before any user-facing reply. Bans jargon, filler and self-coined terms; enforces conclusion-first short output.

Research project skeleton and file lifecycle. Load when initializing a `.kilo/` project, migrating or archiving documents, or deciding where to put a new research note. Defines the four-directory layout (proposal/project/archive/knowledge), two-digit numeric prefixes, the global.md index layer, archive SUMMARY.md rules, and final REPORT.md requirements. It does not run experiments (experiment_manager) or judge whether an idea is worth pursuing (research_progress).

Research idea convergence advisor. Load when the user proposes a new research direction, asks "is this idea worth pursuing", requests literature search / gap analysis, or needs stop criteria for a study. Grounds ideas in reality — implementation closure, related-work facts, speculation tracking, decision-driven experiments — before any formal experiment design. Experiment execution itself belongs to experiment_manager.

Evidence interpretation, baseline-expected-actual comparison, mechanism diagnosis, and statistical verdicts. Load when the user asks to analyze results, interpret data, decide whether to kill/pivot/continue a direction, or evaluate numeric evidence. Locates which causal link failed, enforces convergence accounting, and delegates chart generation to result_visualization.

Charts, evidence tables, process/argument diagrams, and figure validation for research reports and direction discussions. Load when producing or reviewing any long report, when a direction discussion needs implementation-chain/risk/decision diagrams, or when the user asks to plot data, draw diagrams, or produce publication-ready figures. Delegates statistical analysis and verdicts to result_analysis.

Skill-family self-improvement inbox. Load when using any skill reveals friction, a user correction worth persisting, a missing scenario or reference, or a cross-skill inconsistency. Records improvement proposals and applies them only after user review. Never silently edits skills.

Self-contained HTML slide decks as a PPT replacement. Load when the user asks to present the project (or a subset such as motivation, contributions, methods, experiment results) as a horizontal-paging HTML file. Scrolling HTML report rendering belongs to write_md.

Markdown readability, visual planning, and report HTML rendering. Load for any long research report (two passes: visual planning before figures, final readability after), when the user asks to improve readability or format a document, or when a report needs a self-contained HTML reading version. Handles language, structure, and a renderer-aware presentation layer. Horizontal slide decks belong to slide_deck.