bookwormsuf
- 11 skills
- 0 followers
- 11 hours ago last updated
- ▌ Link Pass · bookwormsuf bundleScan recently changed second-brain notes for entity mentions that should be [[wikilinks]], propose them for approval, and flag repeatedly mentioned entities that have no note yet. Use when the user says '/link-pass', 'run a link pass', 'fix my wikilinks', 'find missing links', 'what notes am I missing', or at the end of a session when notes were written or edited and the user wants the vault's links kept healthy. Also suggest it after /review-session when new notes mention existing people or projects as plain text.
- ▌ Explore UI · bookwormsufGenerate multiple interactive UI options for a feature so you can compare and pick the best approach before implementing. Use when the user says '/explore-ui', 'explore options for', 'show me different ways to', 'prototype options', 'compare UI approaches', 'what are the options for this UI', or wants to see multiple interaction patterns side-by-side before committing to one. Also use when the user is unsure which UI pattern fits best and wants to see alternatives. Do NOT use for building a single known component (use /prototype instead) or for non-UI work.
- ▌ Research Code · bookwormsuf bundleCode interview transcripts for qualitative research. Apply codes from a codebook, flag unclear audio, mark quotable moments. Use when: '/code P3', '/code-all', 'code the transcript', 'apply codes to P4', 'code all transcripts in parallel', or user wants to analyze interview data against a codebook. For multiple transcripts (3+), suggest parallel mode for faster results. Also handles UT (usability testing) transcripts with issue extraction and mental model coding.
- ▌ Research Coverage · bookwormsuf bundleCheck which of a study's research questions or session phases actually have coded evidence behind them, and flag the ones that have none. Use this whenever the user asks what they are missing, whether they have enough data, whether they can stop recruiting, if a question went unanswered, "did we answer everything", "what gaps are left", "is this study done", or wants to know whether a finding rests on a single participant. Also use it before writing a synthesis, before a research handover or readout, and when deciding whether another round of sessions is needed. Works with UT studies (checks session phases) and interview studies (checks research questions) that follow the participants/P#/transcript.md convention.
- ▌ Transcript To CSV · bookwormsuf bundleConvert interview/UX research transcripts into CSV files for qualitative coding analysis. Use when the user has a transcript with SPEAKER_XX: utterance format (markdown or plain text) and wants to create a spreadsheet for coding, tagging, or analysis. Triggers on: 'convert transcript to csv', 'parse transcript', 'create coding sheet from transcript', 'transcript for qualitative analysis', interview transcripts, UX research sessions.
- ▌ Fable Orchestrator · bookwormsufDelegation playbook for running multi-phase work as a top-tier model (Fable) orchestrating cheaper tiers. Use whenever a task spans multiple phases or many files — building an app or prototype from scratch, a multi-step feature, a codebase migration, or a research project with many sources. Also use when the user says "orchestrate this", "build this app", "use the cheaper models", or mentions token efficiency on a big task. Not needed for single-file edits or quick questions.
- ▌ Capture Screenshots · bookwormsufCapture screenshots of a running prototype or web app using Playwright MCP, then save locally, upload to Figma, or embed in a markdown file. Use when the user says 'take screenshots', 'capture screenshots', 'screenshot the prototype', 'capture sprint', 'visual changelog', 'screenshot this for the PRD', 'record what the UI looks like', or mentions documenting prototype state for UTs, portfolio, PRDs, or design review. Also use when the user finishes a sprint or feature and wants to save the current state of the UI. Covers any web app on localhost or a URL.
- ▌ Research Transcribe · bookwormsufConvert raw interview transcripts into formatted markdown for qualitative research. Creates participant folder, converts to table format with row numbers, and sets up for coding. Use when: 'transcribe P3', 'add transcript for P2', 'process this interview', or user pastes raw transcript text with SPEAKER_XX format.
- ▌ Research Counter Bias · bookwormsuf bundleStress-test qualitative research findings by hunting for evidence that contradicts them, using fresh subagents that read blinded transcripts so they cannot anchor on the existing codes. Use this whenever the user wants to challenge, pressure-test, sanity-check, or find holes in research themes, findings, an issue log, or a synthesis; whenever they ask "is this finding real", "does the data actually support this", "what am I missing", "am I seeing what I want to see", or "how confident should I be in this theme"; and whenever they are about to present or hand over research conclusions and want them checked first. Also use it to verify frequency claims like "4 of 5 participants" against the transcripts. Works with UT studies and interview studies that follow the participants/P#/transcript.md convention.
- ▌ Landing Page Prototype · bookwormsuf bundleBuild high-quality HTML landing page prototypes by forensically studying a reference website's design system and adapting it for a brand brief. Use when the user provides a reference URL and wants a landing page prototype, or says things like "build a prototype based on [URL]", "create a landing page inspired by [site]", "prototype this for [product]", "make a landing page like [reference]". Also use when the user mentions brand exploration, brand expression prototypes, or landing page variations.
- ▌ Mobbin Reference Sweep · bookwormsuf bundleCollect design references on Mobbin using a three-sweep method that produces genuinely different directions instead of three versions of the same idea. Use this whenever someone wants design references, UI examples, prior art, or inspiration — including 'find references for X', 'how do other apps handle X', 'search Mobbin for X', 'show me examples of X', 'what are the options for this UI', or when they're exploring approaches to a design problem before committing to one. Use it even when Mobbin is never mentioned by name, and even when the request sounds like a simple lookup, because the value is in searching three different ways rather than accepting the first obvious answer.