Skills Claude already knows how to use.

Remove AI writing patterns from text: hard bans (em dashes, Claude-isms, negative parallelism, bold-first bullets), frequency violations (rule of three, rhetorical questions, repeated punchy fragments) and voice problems (corporate jargon, inflated stakes, sycophantic tone, vague attributions), through a tiered audit: hard ban pass, frequency check, voice calibration, self-critique, final. Use when someone asks for it directly: "humanize this", "make it sound less like AI", "de-AI this", "does this read as AI?", "make this sound like a person wrote it", "clean up the AI tells", or when someone asks for a final pass on text before it goes to a real person. Not for flagging AI writing in a document without changing it (a separate skill), for brand or visual style, or for writing new content from scratch.

Build WWTDigital presentations, slides, decks, and branded visual documents using the Figma-derived WWT design system v4.13, and validate any output against it. Use when the user asks for a WWT, WWTDigital or WWT Digital slide, deck, presentation, pitch, divider, cover, or branded HTML/PPTX visual; when they say "match our template", "use the WWT brand", "apply the design system", or "make it look like WWT"; when composing a custom WWT slide layout no template covers; or when checking a WWT deck for design-system compliance. Consult this before writing any WWT-branded markup. Not for other brands, generic slide advice, or documents that are not visual (use the presentation skills for copy).

v4.13.0Figma included Bundle

Check what this machine can actually enforce for the WWT design system, and report whether a deck built here will be validated, lint-checked only, or unchecked. Use when starting any WWT deck (run it before building); when the user asks whether the validator will run, why their output differs from someone else's, or what they need installed, including whether the Aptos fonts are on the machine; when a WWT script fails to launch; and when handing a deck over and needing to state honestly how much of it was checked. Not for checking a finished deck's design (that is the design system's validator).

Publish an HTML page, report or prototype as a shareable, access-controlled link through the team's artifact publisher (the artifact-publisher MCP server bundled with this plugin), and manage who can open it: password gate, expiry, per-person share links, revocation. Use when someone says "publish this", "make this a shareable link", "send the client a link to this report", "put this page somewhere people can open it", "who has access to this page" or "revoke that share". Not for Claude.ai Artifacts (use the Artifact tool), for deploying apps or anything with a backend, or for anything that isn't a single self-contained HTML or Markdown page.

Build a supplemental intelligence brief for a WWT Digital pursuit at any stage: RFP, RFI, or no solicitation yet. Reads the pursuit's Slack channel, Notion space and SharePoint folder (RFP or RFI, working deck, Q&A doc), sizes the deal against comparable past Digital SOWs, checks every claim in our material against wwt.com and the internal record, and researches the client from public sources. Produces a published brief with only what is NEW, the positioning angles the findings support, corrections to the team's own material, and the claims that will blow up if the client checks them. Use whenever someone asks to "look at the pursuit channel," "get up to speed on [client]," "do deep research on [client/project]," "what's new for the team," "size this proposal," "check our credentials," "cross-check the deck," "what's our angle," or asks for a pursuit briefing, deal review prep, or Q&A prep. Trigger even if they only name the client or channel. Not for writing the proposal itself, the SOW, or the estimate.

v1.2.0Needs Slack Needs Notion Needs Microsoft 365 .skill

Run or read a brandscanner assessment of a company's digital presence: security posture, tech stack, app-store reviews, public financials, email-auth records, benchmarks against the scanned cohort, and the resulting scorecard, using the brandscanner MCP server bundled with this plugin. Use when someone says "scan this brand", "run brandscanner on <company>", "what's <company>'s scorecard", "how does <company> compare to its cohort", "pull the tech stack for <company>", or "prep a digital assessment of a prospect". Not for general desk research without the tool, for writing the pitch deck or proposal that follows, or for anything about WWT's own internal systems.

Turn WWT's MyTime timecard emails into Slack nudges from a people manager to the teammates who need to act: the pay period review ("Action Required – Complete Your MyTime Manager Pay Period Review"), missed punch alerts, and past-timecard IDLE Time notices. Sends each person one DM about their own items and the manager one summary of what's waiting on them (approvals, historical corrections). Sets up an hourly scheduled check or runs once. Use when a manager says "set up MyTime nudges for my team", "nudge everyone flagged in today's MyTime email", "ping my team about their missed punches", "chase my team about their timecards", "automate my MyTime reminders", or asks what to do about one of these emails. Needs Slack and Microsoft 365 (Outlook). Not for filling in your own timesheet or reconstructing a Kronos or MyTime week, payroll or time-entry policy questions (use wwtdigital-onboarding), or approving timecards or time off in MyTime itself.

v0.3.0Needs Slack Needs Microsoft 365 .skill

Collapse MyTime's time-off request emails ("Action Required: PTO Request Submitted by <Name>", often sent several times for one request) into one Slack self-DM digest for a people manager: each new request once, cancellations, requests that need manual approval because the day has already passed, other teammates out on the same dates, and a link to approve in MyTime. Sets up a daily weekday digest, or runs once. Use when a manager says "set up a daily digest of the PTO requests waiting on me", "what time off is waiting for my approval", "summarize my team's PTO requests", "too many MyTime time off emails" or "who on my team is out next week". Needs Slack and Microsoft 365 (Outlook). Not for approving or rejecting requests (that stays in MyTime), submitting your own time off, nudging teammates about timecards (use wwtdigital-mytime-nudge), or PTO policy questions (use wwtdigital-onboarding).

v0.1.0Needs Slack Needs Microsoft 365 .skill

Answer WWTDigital new-hire and early-tenure onboarding questions — IT/equipment setup, tool access, benefits/payroll/time entry, PTO, expense reports, team structure, culture/ways-of-working, the 30/60/90-day plan, Digital-specific acronyms, team-specific reference by discipline, and learning & development — from the live Onboarding articles in Notion and their SharePoint sources. Use this whenever someone asks an onboarding-style question about WWTDigital (or "WWT Digital"), such as "how do I get my laptop", "when's my first paycheck", "how do I submit an expense report", "what should I expect on day one", "what does SSA/DPM/NERF mean", "what does my team use for X", or "where do I find the onboarding checklist" (even without saying "onboarding"). Also use to check what's covered before writing new onboarding content, or to audit articles for staleness. Not for a specific person's onboarding board, client-specific onboarding, or ops/pursuit terminology.

v1.1.1Needs Notion .skill

Confirm the WWTDigital skills marketplace is installed and working end to end: the plugin loaded, its skills are visible, and the installed versions are current. Use when someone says "test the skills marketplace", "is the wwtdigital marketplace working", "smoke test the plugin install", or "did my skills install correctly". Reports which install path is in use (plugin or standalone .skill) and the installed skill and plugin versions, by reading its own files; there is nothing to run and no Python needed. Not for validating a skill you are writing; use wwtdigital-skill-author for that.

Author a new Claude skill for the WWTDigital skills marketplace, or get an existing one ready to submit, with or without a git checkout of the repo. Use when someone says "make this a skill", "add a skill to the marketplace", "turn this workflow into a skill", "package this for the team", or "how do I contribute a skill". Produces a SKILL.md in the WWTDigital format (description with trigger phrases, metadata block, numbered steps, a connector check, a verification step), then either runs the marketplace validator in a checkout or applies its rules by hand as a checklist, and drafts the PR or the zip-and-send handoff with test prompts. Not for skills that live outside this marketplace, and not for checking whether an install is working; use marketplace-smoke-test for that.

Implement Figma designs as code by driving a real browser with playwright-cli: extract tokens and measurements from the Figma UI's DOM (not pixels), implement against the token layer, then verify with computed styles plus screenshots at multiple widths. Use when asked to implement or port a Figma design, build a page from a mockup, match a UI to a Figma file, extract design tokens (colors, type scale, radii, spacing) from Figma, or visually verify a UI against a design reference. Not for writing into Figma or generating Figma designs from code (use the Figma plugin's figma-use / figma-generate-design skills), and not for WWT slide decks (use wwtdigital-deck-design).

Build a spec-lite, dependency-ordered implementation plan for a software project or large feature. Use when the user wants to plan a project, scope a body of work into shippable units, write an implementation or roadmap doc, break an epic into stories, or set up a repeatable planning framework. Produces a Waves→Stories plan with verifiable acceptance criteria mapped to tests, anchored to a reality inventory, plus a reusable sequencing-principles doctrine. Honours a repo's .claude/wave-conventions.md for plan paths and story IDs. Not for reviewing a wave that has already been built (use wave-review), and not for writing a single ticket or a status report.

Run an end-of-wave checkpoint on a Waves→Stories implementation plan. Use when a wave has finished, or before starting the next one, to verify each story's acceptance criteria actually passed, update the plan document's status markers and AC checkboxes, detect drift between plan and reality, re-verify the next wave's dependencies, and capture new sequencing lessons. Honours a repo's .claude/wave-conventions.md. Companion to wave-planning. Not for creating a new plan (use wave-planning) and not for reviewing a single pull request's code.