claude-skills

Runs a beginner-mind end-to-end UI audit of any running app — local dev server, staging, production, or a specific URL. Drives Chrome through every interactive element on the target surface, collects structured findings (severity, category, where, symptom, impact, repro, triage), and hands the result off to `/triage-board` which produces the Desktop folder (schema + JSON + Markdown + single-file HTML viewer with MD/CSV/JSON exports and a per-finding Copy as Markdown button). Use when you want fresh-eyes verification of a feature, page, modal, flow, branch, or whole app — before shipping, before review, before a demo, or any time the UI deserves a careful poke.

>-

Reviews the user's past Claude Code conversations from a wellbeing perspective — sentiment, tone, emotional arc, recurring patterns — and generates a supportive, science-grounded report in both Markdown and HTML. Default lookback is 48 hours across all projects. Uses recognised emotion frameworks (Plutchik, Ekman, Russell's circumplex, Pennebaker linguistic markers) and cites the science behind every observation. Learns the user's baseline tone over time so future reports flag genuine shifts, not noise. Use when the user asks for an emotional/wellbeing recap, mood check, sentiment review, or wants to understand their own ups and downs across recent work sessions.

>-

>-

Generates artistic SVGs directly as code — minimal line icons, geometric marks, generative patterns, or hand-drawn compositions — and assembles a modern 2026 HTML gallery preview. Saves each session to its own semantically named folder on the Desktop. Learns from feedback: which styles you loved, which compositions you rejected, which stroke weights and palettes you return to. The tone and aesthetic get more personalized each session. Use when you want icons for a product, a branded mark, a generative poster, or a set of decorative SVGs for a page, blog, or app.

Audits a repository's documentation for unused docs, wrong details, missing coverage, inaccurate data, broken structure, and writing best-practices. Generates a markdown report (HTML optional) with severity ratings. Supports modes (main, comprehensive, focused, quick), per-profile templates (open-source, internal-docs, blog, nextjs-app), optional --fix for low-risk auto-fixes (broken links, frontmatter, stale dates), and a /feedback subcommand that promotes recurring signals into Learned defaults. Use when you want a "docs doctor" pass on a repo before a release, after a refactor, or as a recurring DX health check.

Plays the role of a real user in an interview-style usability test. Takes context about the product (URL, description, focus areas), auto-suggests a persona (overridable), then drives real Chrome via the claude-in-chrome extension to explore the product think-aloud style — narrating live reactions while flagging confusion, friction, and delight, and capturing GIFs/screenshots of the friction moments. Produces a structured UX report with severity-ranked pain points, what worked, and prioritized recommendations. Learns over time which findings the user considers signal vs noise. Use when you want a real-user-perspective audit of a web product to catch UX/UI/product issues that automated tests can't catch.

Runs UI tests described in plain English by driving real Chrome via the Claude-in-Chrome extension. Covers end-to-end flows (clicks, forms, assertions), visual checks (screenshot + optional baseline diff), accessibility (axe-core), performance (Web Vitals + light Lighthouse-style metrics), and an interactive --debug mode that tails console + network and surfaces issues without needing explicit assertions. Accepts inline descriptions or test files in `./tests/ui/*.md`. Per-run folders contain artifacts (screenshots, console, network, axe report, perf metrics) plus a single-file 2026-aesthetic HTML report with a verdict-forward layout. Learns from per-run feedback — false-positive categories, screenshot strategy, severity floors, and per-project baselines personalize future runs. Use when the user wants to verify a UI flow, screenshot a regression, audit accessibility, profile a page, or debug something visibly broken in a real browser session.

Fetches unresolved PR comments, categorizes them (must-fix, suggestion, question, nit), proposes fixes or replies for each, and executes approved actions. Use when addressing PR review feedback or when someone requests changes on your PR.

>-

Generate a stakeholder-aligned next-steps checklist for a multi-stakeholder project from the current conversation and connected context. Use when the user has just had a working session and now needs to align teammates (PM, design, engineering, leadership) on what happens next, who owns it, and what's blocked. Output is grouped by owner, prioritised, and pasteable into Slack/Notion/Linear.

>-

Generates brief, truth-based release notes for a release PR (typically main → prod) by listing each squash-merged PR, a one-paragraph summary drawn from each PR body, and the linked issue-tracker tickets (Linear, Jira, GitHub Issues, Asana, ClickUp, Shortcut, Plane, etc. — any tracker, configured via a URL template). Use when opening or updating a release PR.

Comprehensively addresses PR feedback by fetching all comments, deploying parallel agents for critical analysis of each suggestion, making implementation decisions, then systematically implementing fixes with incremental commits and browser validation. Use when you need to systematically process and implement all PR review feedback with due diligence validation.

Automatically navigates a web app using Playwright MCP and captures context-aware named screenshots at each product feature state. Names each file semantically based on context (e.g., checkout-payment-form-filled.png). Outputs a manifest.json mapping filenames to descriptions and a summary report. Use when documenting product features, generating demo screenshots, building user guides, or creating visual test assets for any web application. Composable primitive — other skills (user-guide, demo-docs) consume its manifest.json output.

>-

Scans the current conversation for patterns worth turning into a reusable skill or memory rule — feedback you've given Claude, workflows you've walked through, corrections, best practices, recurring setups. Proposes candidates, classifies each as skill (multi-step workflow) or memory (single rule), cross-checks against existing skills + MEMORY.md to avoid duplicates, then hands off to /skill-creator (for skills) or writes the memory entry directly. Use when the conversation contains reusable knowledge you don't want to teach Claude again — "we just figured this out, capture it", "turn this into a skill", "save this workflow", "from experience to skill".

Cleans terminal-formatted text and copies it to the macOS clipboard via pbcopy, formatted for the target platform (Gmail, Slack, LinkedIn, plain, markdown). Strips uniform leading indentation, trailing whitespace, and soft-wrapped paragraphs that come from copying out of a terminal, then applies platform conventions (Slack `*bold*` vs Gmail plain prose, etc.). Use when the user wants to paste a draft from the chat into Gmail, Slack, LinkedIn, or any other tool without nightmare terminal formatting (extra whitespace, hard wraps, indented blocks) coming along.

>-

Helps the user decide on a purchase by taking the product links they're considering, asking a couple of sharp clarifying questions about their needs and circumstances, then opening each link in real Chrome via the Claude-in-Chrome extension to extract price, specs, ratings, reviews, return policy, and shipping. Produces a modern 2026 HTML report with a side-by-side comparison, pros/cons per option, an external-review pulse-check, and a clear verdict (buy / wait / pick X over Y / skip). Saves each decision session to its own folder. Learns from thumbs-up/down feedback — brand, budget style, deal-breakers, and past regrets personalize future verdicts. Use when the user pastes one or more product URLs and wants a second opinion before buying, wants a comparison between links they already shortlisted, or asks "should I buy this / which one should I get?".

Audits a website's SEO discoverability by cross-referencing the codebase with live production. Catches sitemap, robots.txt, and canonical inconsistencies, framework-specific anti-patterns (Next.js App Router metadata cascade, Astro/Nuxt equivalents), redirect-chain issues, and Google Search Console stuck states. Produces a prioritized, evidence-backed report with suggested fixes. Use when Search Console shows "Couldn't fetch", pages aren't being indexed, or you want an SEO health check before a launch or after a routing refactor.

Researches products across Amazon, Google Shopping, and relevant specialty sites using the Claude-in-Chrome browser extension. Finds candidates matching user-specified criteria (gift, gadget, gear, home goods, etc.), captures screenshots and key data per candidate, then generates a modern 2026 HTML report with pros/cons, review highlights, metrics, and recommendations. Saves each research session to its own semantically-named folder. Learns from user feedback over time — preferences, favored brands, budget style, and thumbs-up/down on past picks personalize future searches. Use when the user wants to shop for something, compare products, find a gift, or research purchases before buying.

>-

Converts an exploration conversation (architecture discussions, codebase analysis, design decisions) into a structured technical specification document. Generates a comprehensive but concise product-engineering doc suitable for PMs, tech leads, designers, frontend/backend engineers, and QA. Use when you've finished exploring and want to formalize decisions into a shareable deliverable.

Reviews the latest git history, branch state, Linear ticket, and open work to build a concise situational summary. Use when picking up work after another agent, resuming a session, or onboarding to a branch mid-flight.

Investigates GitHub Actions workflow failures for any repo. Fetches recent runs, identifies failures, extracts error logs, diagnoses root causes, and suggests fixes. Use when a deploy or CI workflow fails and you need to understand why.

Generates a contextual Slack message for posting a PR to the team's review channel. Pulls context from PR diff, Linear ticket, session conversation, and related PRs to write a concise, informative review request. Use when sharing a PR for review, posting to the review channel, or preparing a review request message.

Analyzes a repository or branch and generates a meaningful, engineer-friendly timeline of changes — grouping commits into logical units with short and detailed descriptions, using git history, changelogs, GitHub PRs, and Linear tickets. Use when you want to understand what changed, when, and why in a codebase.

Crafts a response to a pasted message (LinkedIn, Slack, Gmail, Teams, etc.) in the user's configured tone and voice. Loads platform-specific context and formatting rules, generates a response matching the platform's conventions, and copies it to clipboard for immediate pasting. Use when you receive a message and need to reply quickly in your own voice.

>-

Checks the current weather for the user's location using live online data. Asks for location on first use and saves it for future runs. Use when the user wants a quick weather check or forecast.

>-

Implements code in progressive, verified increments — auto-detects the project's toolchain, builds each unit, runs checks (typecheck, lint, test), fixes errors, and commits with semantic messages. Use when building features, implementing milestones, or making multi-step changes.

Audits codebases against compliance frameworks (SOC2, HIPAA, PCI-DSS, GDPR, ISO27001, etc.) using parallel agents per subdirectory/sub-repo. Produces a detailed markdown report with line-level code references. Use when you need to check a directory or monorepo for compliance violations before an audit or review.

>-

>-

Identifies and removes stale git branches, orphaned remote branches, and unused worktrees. Cross-references with Linear (or other integrations) to check issue status before deleting. Use when your repos have accumulated stale branches and you want to tidy up.

Generates polished markdown guides from a directory of screenshots and a narrative. Visually reads each image, filters out redundant or irrelevant captures, organizes them contextually, and produces a Notion-compatible markdown file with image placeholders and structured sections. Use when you have screenshots and want to create a product guide, demo walkthrough, or tool guide.

Scans a folder for recent screenshots, visually classifies which ones are relevant to current work, and organizes them into a target directory with descriptive filenames. Use when collecting screenshots for PRs, bug reports, docs, or Linear issues.