Research & Data

Use when the user asks to ingest a book or book chapter into the scientific wiki. Triggers on: "/wiki-ingest-book [title]" (book overview), "/wiki-ingest-book [title] > Chapter N" (chapter-level ingest), or any request to "add a book to the wiki", "ingest this textbook", "add chapter N of [book]".

Diagnostic and repair toolkit for a scientific markdown wiki (Obsidian-compatible): 33 health checks over links, provenance, frontmatter, tables, encoding, synthesis blocks and seals.

Keep a live, deterministic progress report for a Markdown plan checklist or a Spec Kit feature. Collects task state from the plan file, hook events from Claude Code or Codex, and lane, blocker and decision facts you record, then renders a browser view and a Markdown export with no agent spent on upkeep.

Persuasion science from cognitive psychology and behavioral economics. Use to write pitches and offers, reframe pricing or choices, handle objections, diagnose why an audience resists, apply framing, anchoring, or social proof, and audit tactics for manipulation.

Delegate a task to an external agent CLI harness — Claude Code, OpenAI Codex, OpenCode, GitHub Copilot, or Pi — and get back a normalised result with status (successful / failed / abandoned), a summary, what actually changed on disk measured against git, and total input and output token usage.

Mapa de calor de tus anuncios y tus landings. (1) VIDEOS de Meta Ads, retención segundo a segundo con video_play_curve_actions — en qué segundo se va la gente, si el gancho funciona, compras y CPA por video. (2) LANDINGS con Microsoft Clarity — scroll, tiempo activo, clics de rabia/muertos y un veredicto simple por página.

Documentation style for measurement-driven research — "how we cracked it" writeups as a single self-contained HTML page with a story timeline, method, confidence-tagged spec (HW/FILE/HYP/OPEN), verification record and open questions. Use whenever the user asks to document a project or research effort, write a writeup / research log / "how we figured it out" doc, summarize a reverse-engineering, debugging, protocol or hardware investigation, or asks for docs "in the TriCaster / waldobjekt style" — a plain "write up what we did" qualifies too.

Extract income statement, balance sheet and cash flow figures from company financial report PDFs (annual and interim/quarterly reports, scans included) into a relational database, with arithmetic checks so only trustworthy figures load. Use this skill whenever the user wants to extract, load, recover or audit financial statement figures from report PDFs, build or extend a financials database from exchange filings, fix reports that are PARTIAL/FAILED, transcribe statement pages from images, run subagent batches over many reports, or check why a statement did not load - even if they only say "do the remaining reports", "do the next years" or "why is X missing". For Colombo Stock Exchange (CSE, Sri Lanka) work, also use the cse-financials skill.

Build and read a House of Quality (QFD) through the hoq MCP server, with a person. Use when asked to run a QFD, a House of Quality, a needs-to-characteristics matrix, to prioritise characteristics against stakeholder needs, or to check what a product build must hit. Needs come from stakeholders in their words with a source, importance is theirs to give, every characteristic has a unit, room cells and roof pairs are proposals the person confirms, gates are never ranked, and the dated TSV is the version of record.

Use when solving, checking, or explaining a mathematical problem, searching the bundled OpenAI math corpus, or verifying Lean formalizations.

Investiga y redacta derecho chileno con citas verificables. Busca en un corpus local de 9 normas y 6.852 textos de artículo (LeyChile/BCN) y devuelve norma, artículo, texto y URL. Úsala cuando pidan "qué dice la ley", "cita el artículo", revisar o redactar contratos chilenos (compraventa, arrendamiento, mandato, poder) o resolver dudas de Código Civil, Código del Trabajo, Código Penal, Código de Comercio, IVA, consumidor, datos personales o notarial.

Review supplied non-DNA material for a specific identity, relationship or date question. Compare evidence, dependence and contradictions, then give a reasoned conclusion or uncertainty report without changing the tree.

Turn a supplied family interview or transcript into source-linked tree proposals and a complete ledger of people, facts, relationships, stories, photographs and corrections. Use for oral-history processing, without independently searching for relatives or writing to a real tree.

Conducts a comprehensive, source-backed competitive analysis for agency clients — from local SMBs (yoga studios, cafes, clinics) to mid-size commercial businesses. Use this skill whenever a user requests competitor research, a market landscape analysis, a competitive audit, information about who else is in a given market, or asks "who are the competitors of X", "what are others in the [industry] space doing", "analyze the competition for my client in [city/niche]", or any variation of "I need to understand the market before I pitch/plan/launch". Always use this skill before building an ICP, marketing plan, or campaign strategy for a client — competitive context should come first. The final deliverable is always a .docx Word file with proper RTL Hebrew formatting.

Run a real deep dive on any topic. Uses the RECON research engine to find and score the best YouTube videos, academic papers, podcasts and X/Twitter threads (plus a forced contrarian source), loads them into a NotebookLM notebook, interrogates it, verifies the key citations, then delivers an opinionated, cited verdict with a plan. Works without RECON in lite mode (web search + direct reading). Trigger on "deep dive", "deep dive on X", "research X", "ultra think", "think the impossible", "/deep-dive", or any request to go deep on a topic before deciding or building.

Guided build of a custom FinOps cost-attribution (cost allocation) system on FOCUS billing exports from AWS, Azure or Google Cloud, shaped to one organization's accounts, tags, clusters and SaaS vendors, in the user's own stack (the reference stack is S3, Athena and Lambda). Scans the user's bill first, interviews only for what the data can't show, records the answers in a validated profile, decides READY / BLOCKED / NOT_A_FIT per component, then builds phase by phase from specs and golden test cases, with a verification gate after each phase. Use when the user wants chargeback or showback, to allocate or split cloud spend by team, app or cost center, to replace or avoid a commercial FinOps or cost tool, to work with FOCUS exports, to split shared costs such as Enterprise Support, to attribute Kubernetes namespaces or SaaS vendor bills, to resume or refresh a build this skill started, or to try it on a sample bill before giving it billing access.

Researches any company in depth from its name or URL and delivers a designed PDF brief styled in that company's own colors and type, covering financials, full history, leadership, hiring stats and pay, employee reviews, a brand kit with links to official vector logos and current brand guidelines, advertising, activations, social, videos, podcasts, news, competitors, risks, and exactly how to talk to them in an interview or pitch. Use when someone asks for a company brief, company research or a deep dive on a company, interview prep for a specific company, whether to work for or accept an offer from a company, how to pitch or approach a company as a client, or brand research before designing for or working with a brand for the first time. Usage - /company-brief followed by a company name or URL, optionally "for" a role, optionally "brand kit", optionally "shareable".

Build a comprehensive Hebrew client profile document (.docx) for a new or prospective agency client. Use this skill whenever a agency employee asks to profile a client, research a potential client, build an "אפיון לקוח", or prepare a client strategy document. Triggers: "אפיין לקוח", "תבנה פרופיל לקוח", "תכין מסמך אפיון", "client profile", "profile this business", "research this client", any request to profile/analyze a business for marketing purposes, or any time a business name/link is shared with the intent of creating a marketing strategy document. All output is in professional Hebrew. The deliverable is always a Word (.docx) file.

Use when auditing financial spreadsheets (.xlsx or .csv), detecting accounting fraud, split transactions, Benford's law anomalies, or internal control defects, retrieving Chinese auditing standards (CSA) or enterprise internal control regulations, or generating CSA-1131 compliant audit working papers (.docx and .md). Triggers on financial audit, voucher review, audit workpaper, split transaction fraud, 财务审计, 凭证核查, 审计底稿, 拆单舞弊, 法规检索, 内控缺陷核查.

Comprehensive website content and UX audit. Use this skill whenever someone asks to review, audit, check, or improve a website, landing page, or web product. Triggers include: "audit my website", "review my site", "check my content", "is my website ready to launch", "what's wrong with my site", "content review", "UX review", "website feedback", "check for broken links", "check for lorem ipsum", "review my copy", or any request that involves evaluating an existing website's content quality, UX structure, copy, or readiness for launch. Also trigger when someone shares a URL and asks for feedback, improvements, or a teardown. This skill combines content editing, UX strategy, creative direction, and technical QA into a single structured audit.

Audit websites, apps, screenshots, mockups, and design code for harmful AI-design clichés, generic generated defaults, and established UI defects. Use when the user wants evidence-backed design feedback, an anti-slop review, or a removal-first cleanup plan without a speculative redesign.

Use when conducting a read-only, reputation-safe Website Data Protection Impact Assessment (DPIA) & Client-Side Automation Readiness Audit under India's Digital Personal Data Protection Act, 2023 (DPDP Act) and DPDP Rules, 2025 (G.S.R. 846(E)). Detects pre-consent cookies, storage, marketing/CRM automations, CDPs, session replay, AI chatbots, ad pixels, dark patterns, security headers, 90-day grievance SLA signals, and cross-border sub-processors across Claude, Gemini, DeepSeek, and Qwen.

Analyses GA4 landing pages for engagement, key events and leaks, paid vs organic fit with Google Ads cost and Search Console clicks, scroll depth, 404 pages and site search. Use when asked which landing pages work, which convert by traffic source, or which pages return 404.

Audits a landing page, homepage or sales page for conversion using GA4, Search Console, Google Ads and desktop and mobile renders, scoring seven pillars and ranking fixes with rewrites. Use when a page isn't converting, or for a landing page review, mobile check or new page structure.