persona-documentarian
Adopt the Documentarian stance: human-facing docs for a reader who hasn't read the code and has one question — one Diátaxis frame throughout, every example run…
Skills
Find an Agent Skill for your task. Read the guide and copy the install command.
Updated 11-06 00:14
Writing & research
14,813 results
persona-researcher
Sharpen depth research into inquiry: claims grounded in checkable primary sources, observation distinct from claim, no decision committed. ALWAYS apply when…
persona-surveyor
Adopt the Surveyor stance for breadth / inventory research — UX, market, and competitive surveys of what prevails across many examples (what competitors do,…
write-documentation
Implement a documentation task: write or update human-facing docs (README, tutorial, how-to, reference, explanation) — one frame held throughout, every example…
research
Run three parallel research tracks after an approved brainstorm doc: codebase patterns, framework/library docs, and prior solutions from substrate. Writes…
long-blog-to-viral-social-posts
Turn a long-form article into three differentiated LinkedIn and X posts with source-grounded hooks, an inferred audience, and first-comment CTAs. Use for blog…
digest
Write a dated digest — a prose synthesis of everything captured and changed in a time window, grouped by domain tag and cross-linked.
ingest
Ingest a source into the wiki — fetch a URL or video, or sweep the raw inbox, then file a durable summary, a raw backstop, and fold it into every page it…
lint-wiki
Health-check the whole wiki — contradictions, stale claims, orphans, gaps — plus where it should grow next, with candidate pages for the user to pick. Report,…
lokalise-install-auth
Install and configure Lokalise SDK/CLI authentication. Use when setting up a new Lokalise integration, configuring API tokens, or initializing Lokalise in your…
lokalise-webhooks-events
Implement Lokalise webhook handling and event processing. Use when setting up webhook endpoints, handling translation events, or building automation based on…
notion-knowledge-capture
Transforms conversations and discussions into structured documentation pages in Notion. Captures insights, decisions, and knowledge from chat context, formats…
notion-research-documentation
Searches across your Notion workspace, synthesizes findings from multiple pages, and creates comprehensive research documentation saved as new Notion pages.…
fantasy-name-generator
Generate fantasy character names in zh-TW with original spelling and etymology, across human, elf, dwarf, orc/beastfolk, and halfling naming traditions. Use…
writing-style-guideline
Language-agnostic writing style guidelines covering structure, tone, rhetorical constraints, banned phrases, and review checks. Use when writing any kind of…
cross-video-qa
Use when the user asks questions that require finding answers across multiple videos, such as "which episode did I talk about X in", "who said Y in these videos", "how is X explained in my videos", or "which videos discuss X": search by meaning, deduplicate by video, re-locate against the current version, answer only from the retrieved original words, and cite the video and timestamp for each claim; state when coverage is incomplete. Not for summarizing or rewriting a single video (that is summary, blog), and do not modify videos.
explainer-video
Use for knowledge explainers, popular science, and teaching videos: take a concept, a news story, a paper, or a piece of material and turn it into an explainer video that someone with no background (default: high school graduate level) can understand, will want to finish, and can retell after watching. What's given here is judgment and technique, not a fixed template: first decide the form of this episode based on the content, find the way to explain it starting from the audience's questions, give the whole video a throughline subject (the thing itself is the explanation), let visuals and narration be driven by the same set of facts, verify the facts, and produce additional… as needed
polish-transcript
Use when proofreading or polishing a video transcript: fix typos and mishearings, unify terminology and proper names, correct punctuation and sentence breaks, and write back to the same transcript with word timings unchanged. Do this before producing a translation or building a subtitle layer. By default, do not change grammar, delete filler words, or condense; removing fillers and pauses is editing and is not done here. Not for polishing articles or plain text unrelated to the video.
talking-head-cut
Use when a recording of one or two people talking to camera needs to be cut into a publishable video: remove verbal tics, long pauses, false starts and retakes, and off-topic remarks; pick one take among several retakes; tighten pauses between sentences; move the strongest original line to the opening; then add punch-ins for emphasis, B-roll, vertical framing, and music ducking as needed. Requires a transcript of the video. Use short-form clipping when you only want to pick a few short segments out of a long video.
titles-and-description
Use when the user wants to title a video that is about to be published, wants several title candidates, wants another batch or a few more modeled on a specific one, or wants a video description (the explanatory text under the title on the video page, chapter list, tags). Written for people publishing videos: be compelling, and let every sentence be delivered by the video. Requires the video's transcript. Not for titling articles or other writing unrelated to video.
translate-subtitles
Use when the user wants to translate a video's transcript or subtitles into another language, make bilingual subtitles, or update translations that have gone stale after changes to the source text: translate sentence by sentence and write the result as the video's translation document, then place it on screen and export as needed. When media is present, "translate" means subtitle translation by default. Do not use it to translate articles, documents, or a plain block of text unrelated to a video; if the conversation has no video or subtitles and it is unclear what is meant, ask once first.
video-blog
Use when the user wants to turn a video into a blog post, a WeChat official account or column article, or an illustrated article: a complete article that can be understood without watching the video, divided into the author's perspective (the user's own video, published in the user's name) and the viewer's perspective (someone else's video, retold with source attribution). Requires the video's transcript. Not for cases where only summary key points are wanted (use video content summary), and not for writing unrelated to the video.
video-summary
Use when the user wants to summarize a video, extract key points, write a content summary, or a meeting-notes-style recap: pick out the two or three things from this episode most worth taking away, start from the questions the reader cares about and explain them thoroughly, and say little or nothing about the rest; also attach a set of key points with video timestamps so readers can jump back to that segment. Suitable for interviews, podcasts, talks, tutorials, reviews, meetings, roundtables, and any video with a transcript. Not for summarizing articles or plain text unrelated to the video, and not for writing blog posts (that is a different kind of output)…
readme
Write documentation based on conversation history and folder contents