← all publishers

songyw2003

@songyw2003 source repo

12 published skills

  1. Idea Spark · songyw2003 bundle
    Generate ONE reviewer-defensible, implementable research idea with a concrete method and falsification plan from a stated research direction. Use when the user asks for a research idea, novelty analysis, bottleneck diagnosis, or paper-shape suggestion. Skip code review, debugging, and unconstrained brainstorming without research context.
    0 installs
  2. Paper2reel · songyw2003 bundle
    Build an interactive HTML viewer that aligns paper2poster output with slide/video deck frames through a sidecar content_alignment.json, without modifying the original poster, PPTX, or blog deliverables.
    0 installs
  3. Paper2video · songyw2003 bundle
    Turn a research paper, a paper2assets package, or an existing PPT deck into a narrated MP4 video. Prefer the shared paper2assets package when present so paper2poster, paper2blog, paper2slides, and paper2video use the same section order and narration. Preserve the advanced deck route by delegating slide authoring to the external `hugohe3/ppt-master` project, then synthesize audio with `skills/paper2poster/scripts/generate_audio.py`, render with `skills/paper2video/scripts/render_video.py`, and burn final subtitles with `skills/paper2video/scripts/add_subtitles.py`.
    0 installs
  4. Scoop Check · songyw2003 bundle
    Check whether a proposed research novelty (given a problem statement and claimed idea/novelty) overlaps with existing published work. TRIGGER when the user asks to verify research novelty, check if an idea is new, compare a proposed contribution to prior art, do a literature/prior-art search for a specific claim, or assess whether a paper idea has been scooped. DO NOT TRIGGER for general literature reviews unrelated to a specific novelty claim, or for writing related-work sections of an already-validated idea.
    0 installs
  5. Paper Search · songyw2003 bundle
    Search papers across arXiv, DBLP, OpenAlex, OpenReview, Semantic Scholar, and Crossref for a given query and year range, using ./scripts/search_papers.py. Use when the user asks to find papers, related work, prior art, or recent publications on a specific topic, especially when they mention a date range or specific venues like NeurIPS, ICLR, or ICML.
    0 installs
  6. Paper2assets · songyw2003 bundle
    paper2assets — paper PDF → reusable assets
    0 installs
  7. Paper2poster · songyw2003 bundle
    Render a pre-extracted paper's structured 9-section spec (`paper_spec.md`) into a single-page HTML academic poster, fit the layout to the page via an iterative measured-fill loop, and export it to print-ready PDF + PNG thumbnail. Requires the upstream `paper2assets` skill to have produced the input `<outdir>/` package (`manifest.json` at the root + an `assets/` folder holding `meta/paper_spec.md`, `meta/text.txt`, `meta/figures.json`, `meta/metadata.json`, `figures/*.png`, `logos/`, `qr/`) first. Use when the user wants an HTML poster, PDF/PNG export, or PPTX from a paper they already have extracted assets for — e.g., "render the poster", "make the poster from this spec", "export poster to PDF", "paper2poster". The three skills paper2assets → paper2poster → html2pptx run in sequence, each invokable on its own.
    0 installs
  8. Feishu Research Docs · songyw2003 bundle
    Create and maintain clear, readable Feishu cloud documents from completed research using lark-cli and the user's Feishu identity. Use when a research report, review, investigation, or analysis should be written to a personal Feishu cloud document, not a Wiki or knowledge-base node, including creating a new document, refreshing an existing research document, or converting a local Markdown draft into a polished Docx cloud document.
    0 installs
  9. Show Reproduced Repo · songyw2003 bundle
    Create or refresh a single, minimal HTML showcase for a previously reproduced repository. Use when Codex needs to present verified reproduction results, images, videos, conclusions, current progress, blockers, and concrete next steps in one self-contained white-background, black-text HTML file for a demo, handoff, progress review, or report.
    0 installs
  10. Configure Project Env · songyw2003 bundle
    Inspect this workstation and configure, repair, or reproduce a Python project environment with a conda-first backend (falling back to uv when conda is unavailable or explicitly required), a project-named environment, reusable package caches, parallel package fetching, GPU-aware PyTorch selection, and a runnable smoke test. Use when Codex is asked to configure dependencies, create an environment, continue environment setup, reproduce a repository, install GPU packages, resolve CUDA or nvcc issues, or run a project demo after setup on this machine.
    0 installs
  11. Git Structured Commit · songyw2003 bundle
    Review working-tree changes and create focused Git commits with Chinese Conventional Commit messages using required type and subject plus optional scope. Use when Codex is asked to commit changes, initialize version control and commit a project, split a large change set into logical commits, prepare local commit history, or validate commit messages. Initialize Git when the intended project has no repository, stage only explicit functional groups, preserve unrelated changes, and never bundle the whole working tree by default.
    0 installs
  12. Plain Language Editor · songyw2003 bundle
    Turn LLM-sounding content into plain, human, low-reading-age copy that a smart teenager could follow, in the spirit of how Richard Feynman explained hard ideas with simple words. Use when text reads as AI-generated, hard to follow, padded, or full of meta-commentary, or when the user wants a draft audited for AI-slop patterns without a rewrite. Goes beyond surface AI-tells: it strips notes-to-the- reader and meta-headings, forces every sentence into clear subject-verb-object, defines or deletes undefined jargon, audits each claim for truth and internal consistency, cuts length hard, and cuts named AI-slop patterns (banned words, colon reveals, fake-profound kickers, weasel attribution) while preserving the writer's real voice. Pairs with the humanizer skill (which handles AI vocabulary and style tells); run this for comprehension, substance, trust, and voice.
    0 installs