Lab

Fetches unresolved PR comments, categorizes them (must-fix, suggestion, question, nit), proposes fixes or replies for each, and executes approved actions.

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.

Audits all personal Claude skills against the SKILLS_GUIDE.md manifest, latest official documentation, and best practices. Reports issues and improvement suggestions.

Implements code in progressive, verified increments — auto-detects the project's toolchain, builds each unit, runs checks, fixes errors, and commits with semantic messages.

Automatically navigates a web app using Playwright MCP and captures context-aware named screenshots at each product feature state. Names each file semantically (e.g., checkout-payment-form-filled.png), outputs a manifest.json mapping filenames to descriptions and a summary report. Composable primitive — other skills consume its manifest.json output.

Analyzes a large pull request, branch, or commit range and proposes a sequence of smaller, dependency-aware, merge-safe PRs following review best practices — schema before API before UI, refactor before feature, one concern per PR. Suggests a plan; the user approves. On approval, creates chunk branches, cherry-picks commits, pushes, opens draft PRs, and links them to the parent Linear issue. The parent branch stays untouched.

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.).

Creates a well-structured pull request with product-focused summary, change highlights, and test steps. Auto-detects base branch and links Linear issues.

Scans the current conversation for patterns worth turning into a reusable skill or memory rule — feedback, workflows, corrections, best practices. Classifies each candidate as skill (multi-step workflow) or memory (single rule), cross-checks against existing skills and MEMORY.md, then hands off to /skill-creator or writes the memory entry directly.

Identifies and removes stale git branches, orphaned remote branches, and unused worktrees. Cross-references with Linear to check issue status before deleting.

Investigates GitHub Actions workflow failures for any repo. Fetches recent runs, identifies failures, extracts error logs, diagnoses root causes, and suggests fixes.

Generates a stakeholder-aligned next-steps checklist for a multi-stakeholder project from the current conversation and connected context (Linear ticket, PR, Slack, Notion). Output is grouped by owner, prioritised, and pasteable into Slack/Notion/Linear.

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.

Generates a contextual Slack message for posting a PR to the team's review channel. Pulls context from PR diff, Linear ticket, and session conversation.

Posts a structured implementation summary comment to a Linear issue — what was built, key decisions, reuse patterns, and how to test.

Drafts Linear project status updates (Done / In Progress / Next / Blockers) for each project the user leads, gathering context from Linear issues, git commits across configured repos, GitHub PRs, Slack mentions, Notion edits, and Gmail/Calendar. Never auto-posts — always shows the draft for review. All site-specific values (repos, project list, Linear org) live in preferences.

Publishes personal Claude skills to a GitHub repository for sharing. Copies skill files, generates a README catalog, commits, and pushes.

Generates brief, truth-based release notes for a release PR by listing each squash-merged PR, a one-paragraph summary drawn from each PR body, and the linked issue-tracker tickets. Tracker-agnostic — works with Linear, Jira, GitHub Issues, Asana, ClickUp, Shortcut, Plane, or any tracker via a configurable URL template.

Researches products across Amazon, Google Shopping, and specialty sites via the Claude-in-Chrome extension; produces a 2026 single-file HTML report with pros/cons, review highlights, and picks. Learns from feedback to personalize future searches.

Takes product URLs under consideration, asks two sharp clarifying questions about needs and dealbreakers, opens each link in real Chrome to extract price, specs, reviews, returns and stock, cross-checks independent reviews and price history, then ships a 2026 HTML report with a clear verdict — buy, wait, pick X over Y, or skip. Learns from thumbs-up/down and past regrets to personalize future calls.

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.

Creates new Claude Code skills interactively by asking contextual questions about purpose, side effects, tools, and workflow. Generates a complete SKILL.md.

Creates a Linear issue from a pasted Slack thread. Parses the conversation, infers title, priority, category, and description, checks for duplicates.

Traces and verifies that something works end-to-end in any environment. Builds a check plan from natural language input, then runs each check reporting pass/fail.

Walks through a long enumerated response (PR review findings, audit list, refactor plan, action items) one item at a time, with a per-item decision: dive deeper, take action, defer, or skip.

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. Learns from feedback across sessions: which styles landed, which compositions were rejected, which stroke weights and palettes you return to.

Merges one branch into another with conflict handling. Stashes work, updates both branches, merges, resolves conflicts, pushes, and restores state.

Parses a pasted thread (Slack, email, GitHub, Teams), analyzes it against current context, and suggests actionable next steps — then executes them with confirmation.

Runs UI tests described in plain English by driving real Chrome via the Claude-in-Chrome extension. Covers end-to-end flows, visual checks (screenshot + optional baseline diff), accessibility (axe-core), performance (Web Vitals + light Lighthouse), and an interactive --debug mode that tails console + network. Per-run folder of artifacts plus a single-file 2026 HTML dashboard report (verdict-forward, bento-grid). Learns from per-run feedback.

Suggests the 3 most impactful next actions based on full developer context — git, Linear, PRs, and current conversation. Prioritizes blockers and momentum.

Analyzes recent Claude Code conversations and local Claude state (skills, settings, memory files, CLAUDE.md), researches the latest Claude Code features and best practices online, and suggests one workflow improvement at a time with reasoning and a concrete action item. Can save accepted suggestions to memory for tracking.

Scans all connected communication channels — Gmail, Slack, Calendar, and any available integrations — then produces a prioritized catchup briefing. Helps triage messages and draft replies.

Researches a topic systematically and produces a structured briefing. Gathers key facts, perspectives, and sources into a clear summary.