Docs & Writing

mthines Bundle

Stabilizes or optimizes Playwright E2E tests on one PR via a local-first loop, then ratifies with a single CI run. Pulls Dash0 spans (`git.pull_request_link`) as the historical baseline, then captures every iteration's evidence locally with `--trace=on` (same OTel exporter, same trace schema). Validation is empirical, not predictive: before commit, every new locator must resolve against source (static grep) or the live app (`locator.count()`); after commit, the fixed test must pass three consecutive local runs before the single push. Modes: `stabilize` (default) heals flaky / failing tests; `optimize` is report-only and ranks slow-action wins by measured ms saved. Refuses `.skip`, `.fixme`, `waitForTimeout`, or any check-weakening edit. Use when a PR has flaky or failing E2E tests or when you want to find slow tests worth tightening. Triggers on "stabilize this PR", "fix flaky e2e", "heal playwright on PR", "ui-e2e is failing", "self-heal e2e", "optimize e2e", "/e2e-pr-stabilizer".

jasperpwang Bundle

Polish, restructure, or translate academic prose into Nature-leaning English using writing-strategy principles, curated Nature/Nature Communications article patterns, and phrase-level support from Academic Phrasebank. Use whenever the user asks to polish a manuscript paragraph, abstract, introduction, results, discussion, conclusion, title, methods section, or Chinese academic draft for publication-quality English. Also covers LaTeX layout/typesetting (排版) fixes — loose or sparse pages, stranded section headings, figures that don't fill the page or split across pages, "Float too large", multi-panel figure arrangement, and Supplementary Information that looks empty — via references/latex-layout.md. Also trigger on general academic/scientific writing requests even without the word "Nature", including academic writing, scientific writing, SCI/paper writing, English manuscript polishing, language editing, proofreading, and Chinese phrasings such as 学术写作、科研写作、论文润色、写paper、SCI写作、英文论文润色、语言润色、润色、改写、学术英语、英文写作.

jasperpwang Bundle

Builds a paper or report from materials using the shared PaperSpine research, motivation, and rationale workflow. (internal /paperspine step)

jasperpwang Bundle

Feishu/Lark document and wiki workflow for reading, editing, structuring, and verifying docs, wiki pages, paper cards, meeting notes, deep dives, images, and drive-native Markdown through the local `lark-cli`. Use when the user provides a feishu.cn/larksuite.com document or wiki URL, asks to read or edit 飞书/Feishu/Lark content, asks to maintain Feishu paper cards or meeting pages, mentions 飞书文档/飞书知识库, or needs authenticated Feishu document updates with hierarchy, image, and post-write verification rules.

jasperpwang Bundle

Canonical platform-neutral delivery standard for source-grounded Paper Cards and compact Paper Abstract Cards in Feishu/Lark, Notion, Obsidian/Markdown, surveys, deep dives, meeting notes, and research trackers. Use whenever the user asks to整理/补全/生成/审核 paper card, 论文卡片, Paper Abstract Card, 论文摘要卡片, literature survey cards, meeting paper cards, or deep-dive paper cards; also use when another research workflow will create or modify either card type.

jasperpwang Bundle

Run the user's daily科研经营复盘 / Daily Research Operating Review and persist it in Feishu/Lark, Notion, or Obsidian/Markdown. Use when the user says they want to start, continue, write, or organize a daily review, daily research review, 每日复盘, 科研复盘, 今日复盘, or asks to create/update a dated review page with reflection, project triage, AI ROI, simulation-ready research direction checks, and tomorrow's three key TODOs.

jasperpwang Skill

Obsidian/Markdown vault workflow for reading, editing, structuring, and verifying notes used by research skills (paper cards, deep dives, surveys, wiki pages, meeting notes). Use when the user gives a vault path or .md note, names Obsidian/WorldModelVault as the destination, or research-doc-workflow routes a durable write to local Markdown. Applies Obsidian-native format deltas so Feishu/Notion syntax does not leak into reader-facing notes.

jasperpwang Bundle

Route creation, editing, migration, and verification of research documents across Feishu/Lark, Notion, and Obsidian/Markdown while preserving content contracts, hierarchy, formulas, images, captions, links, and source provenance. Use whenever a task-specific research skill writes or normalizes paper cards, deep dives, surveys, experiment reports, bilingual source archives, daily reviews, meeting notes, or wiki pages, or whenever the user names Feishu, Notion, Obsidian, a wiki/page URL, or a local Markdown destination. Routes to feishu-doc-workflow, notion-doc-workflow, or obsidian-doc-workflow.

jasperpwang Bundle

Canonical platform-neutral source archive workflow for web articles, essays, research-method posts, pasted source text, browser/Obsidian web-clipper captures, and similar non-paper sources delivered to Feishu/Lark, Notion, or Obsidian/Markdown when the user asks for 原文, 提取原文, 英文原文和中文译文, English original plus Chinese translation, bilingual source archive, or says not to summarize. Use especially when another workflow would otherwise summarize or distill the source.

jasperpwang Bundle

Write source-grounded research experiment reports and experiment-log entries for ML/vision/graphics/robotics projects. Default durable destination is Notion; Feishu/Lark or Obsidian/Markdown only when the user explicitly names them. Use when the user asks to整理/撰写/补全/审查实验报告, 实验记录, 训练报告, 评测报告, ablation report, baseline report, reproduction report, Notion experiment pages, or research-document experiment summaries.

jasperpwang Skill

Import a paper's original and pdf2zh Chinese PDFs into Notion as a correctly nested child-page package, then repair the Chinese manuscript against the deep-dive standard. Use when the user gives an arXiv PDF URL or paper title and asks to pdf2zh, import the Chinese paper into Notion, or place translated paper PDFs in a Notion page.

arisng Bundle

Create and evaluate documentation using the Diátaxis framework. Use when writing, organizing, or auditing documentation to ensure it serves distinct user needs through four systematic categories (Tutorials, How-to Guides, Reference, Explanation). Ideal for diagnosing documentation problems, separating mixed content, and ensuring each piece serves a single, clear purpose.

arisng Bundle

Create and draft markdown-based issue documents (bug reports, feature plans, RFCs, ADRs, tasks, retrospectives) in the top-level `.issues/` folder. Use this skill whenever you need to document software issues, feature proposals, architectural decisions, work items, or post-mortems. Includes templates, metadata indexing, and structured YAML frontmatter. Different from issue tracker systems — this is for archival, decision-making, and knowledge base documents.

arisng Skill

Classify documentation files into domain-based sub-categories within Diátaxis top-level categories (tutorials, how-to, reference, explanation). Use when organizing wiki files into sub-folders like ralph/, copilot/, sdk/ within a Diátaxis category, determining where a file belongs using the three-rule heuristic (keyword extraction → reuse check → create check → fallback), reclassifying research/ staging files into standard categories, or performing batch wiki reorganization. Supplements the `diataxis` skill which handles top-level classification — this skill adds sub-category depth.

arisng Skill

Multi-iteration parallel subagent orchestrator for Kimi Code CLI with streamlined observability, automated documentation, and atomic commits. Use when orchestrating complex work across multiple subagents, enabling parallel execution, or when explicitly requesting fleet mode with '/flow:fleet'. Integrates diataxis documentation and git-atomic-commit workflow.

blackbelttechnology Skill

Set up, initialize, or configure the local markdown knowledge base for a project. Use when the user says "set up the knowledge base", "init the KB", "configure the markdown KB", "index my docs", "add a doc source to the KB", or wants the project's markdown searchable by `kb search`. Wraps `kb init` end-to-end through `kb index` and a smoke `kb search`.

blackbelttechnology Skill

Search the local markdown knowledge base for facts, prior decisions, definitions, people, error codes, config flags, and how-we-fixed-it notes. ALWAYS search here BEFORE answering a project-specific question from memory, guessing, or asking the user. Use whenever you hit an unknown term, an unfamiliar entity, an error string, or a "how do we do X / why did we choose Y" question.

blackbelttechnology Bundle

Turn a pi session into a Markdown "how-we-did-it" collaboration guideline: reads the session's JSONL transcript and synthesizes a reusable playbook of which prompts worked, what had to be steered, and how to reproduce the result faster. Use when: "document this session", "write up how we did X with the AI", "make a guideline from this session", "turn this session into a playbook/tutorial".

blackbelttechnology Bundle

Sweep all active OpenSpec proposals for staleness, conflicts, and obsolescence against the current codebase and archived changes. Use when proposals may be outdated, when checking cross-proposal conflicts, or before starting a batch of implementations. Produces a gap-analysis report, updates a priority queue file, and can auto-fix trivial issues or guide conversations for complex ones.

xyva-yuangui Bundle

PDF 文档处理。支持提取文本、提取表格、合并 PDF、拆分 PDF、PDF 转图片、获取元数据。用户需要处理 PDF 文件时使用此技能。

nvidia Bundle

Run a scheduled web-surveillance sweep over a watchlist of topics using Tavily web_search, optional Tavily extraction, deterministic dedup/exclusion filtering, and a cited Markdown digest. Use when the user asks to run a sweep, run a watchtower sweep, monitor the watchlist, check the watchlist, or asks what changed since the last run. Trigger keywords - run a sweep, watchtower sweep, monitor watchlist, what changed.

nvidia Skill

Run configured native OVPhysX demos and report host GIF/status receipts when Hermes is sandboxed but the OVRTX/OVPhysX runtime and visible Blender process are on the host.

3dcom2711 Skill

Publish a hunt as a case report, escalation, detection promotion, or leadership summary

3dcom2711 Skill

Generate a session report with token usage estimates, work summary, and outcomes

magnus919 Bundle

arc42 template — constraints and context documentation covering sections 1-4 (stakeholders, quality goals, constraints, solution strategy) and supplemental sections (architecture, risks, deployment, glossary). Maps arc42 sections to Artifact Pyramid layers (Canvas→L1, sections 1-4→L2, sections 5-12→L3). Use when the technical-architect profile needs to document system context and constraints.

digitalspeed Skill

Read data from a Google Sheet and create a formatted Google Docs report.

jasonxzwen Bundle

Load when beginning every repository mutation task to run one dependency-layered batch interview; a fully aligned task takes the zero-question path, while durable documentation work uses grill-with-docs to reuse the same decision graph.

jasonxzwen Bundle

Load when a task needs a deep, report-only review across independent Standards and Spec axes, using bounded read-only native Subagents when available and adding risk lenses only when the diff warrants them.

jasonxzwen Bundle

Load when turning an external article, blog, release note, interview, or report into a source-backed public post with Chinese fidelity, media references, an effective-interact summary layer, and project-iteration review; do not load for ordinary summaries, full copyrighted reposts, or production site design.

jasonxzwen Bundle

Load when existing animation or motion code, a motion diff, or a rendered interaction needs an evidence-backed craft review; report findings only and use code-review for general source review.

datus-ai Bundle

Wrap a user-provided SDK / REST API / documentation into an installable Datus plugin. Produces a design draft FIRST (config schema, CLI command list with doc citations + permission rules, bundled-skill plan), then STOPS for user confirmation before writing any code. Fully self-contained — the entire datus-plugin.yml manifest contract is inlined below.

jperrello Bundle

Composite a closing CTA banner ("FOLLOW FOR MORE" + @C0BALT_CUT) over the last ~2.5s of a finished short so it lands on an intentional beat instead of dead-stopping on a dangling word. Timeline-preserving (audio copied, duration identical), deterministic, non-fatal. END_CARD=0 skips.

jperrello Bundle

Manufacture "multi-cam" hard cuts on static talking-head stretches by alternating a base framing with tighter punch-in reframings of the SAME speaker, each cut snapped to a word start so it lands on speech. Deterministic, no Claude. Timeline-preserving — audio is copied untouched and total duration is identical, so every downstream timestamp (captions, b-roll windows) stays valid. Runs on the 1080x1920 vertical clip AFTER fill-vertical and BEFORE zoom-punch/broll/burn-subtitles, so cutaways override their own windows and captions burn on top. The visual-change-density lever for sources where b-roll stays sparse.

jperrello Bundle

Semantic dead-air / filler removal. Claude reads a clip-local word-timed transcript and marks filler words, trail-offs, false starts, repeated re-starts, and short digressive asides for removal. Emits keeps.json (ranges to keep) and transcript.trimmed.json (kept words with shifted timestamps). Pairs with cut-filler, which applies the cuts to the clip's video.

jperrello Bundle

Remove dead air from a clip by collapsing any inter-word gap > gap_max (default 0.18s). Gaps collapse to collapse_to (default 0.08s), or sentence_beat (default 0.15s) when the preceding word ends with . ? or !. Re-emits the clip and a re-timed transcript so downstream subtitle / chunk / title skills align. Use right after cut-clip + rebase to keep pacing punchy.

jperrello Bundle

Pick N clip-worthy spans from a transcript. Claude reads the full transcript plus a per-second RMS energy summary and returns the most engaging time ranges for shorts. The atomic unit of a short is ONE IDEA, not one topic — each span carries `cuts` (1-3 non-contiguous source ranges assembled into one flowing story) plus `idea` (the one-sentence throughline all cuts share). A multi-cut span may pull cuts from DIFFERENT topics when they develop the same single idea; a single-cut span must stay within one topic and is just `cuts:[[t0,t1]]`. A 1-second SPOKEN hook is mandatory (no title card rescues a slow open — picks under PICK_HOOK_FLOOR are dropped). Replaces the old keyword/hook-word heuristic ranker.