Claude Skills by calebnewtonusc

Finish the parts of this kit's install that need a browser or a permission dialog: screen and accessibility grants, provider keys, any step behind a web console, and verification. Use after setup.sh, when doctor.sh reports a tool installed but unusable, or when the user says to finish setting something up themselves.

Answer from the A2A Spring 2026 course, 108 hours of workshops on building a business out of AI agents. Use when the user asks what the course said about something, references a week or module by number, asks how Callan or Carter taught a thing, or asks for the course's take on prompting, agent architecture, sales, marketing, operations, finance or building an AI employee. Also use when writing a new Chewbacca skill, to check whether the course already solved that problem.

Turn a piece of work into something the user can listen to instead of read. Use when they say they want to hear it, listen to it, put it in their ears, or that they will be driving, walking, at the gym, doing laundry, or otherwise away from the screen. Also use when handing over a long analysis they will not sit and read, and when they ask for a voice memo, an audio summary, or a recording of a briefing.

Debrief a sales or client call from its transcript or recording. Use when the user shares a call transcript, mentions a call that went well or badly, asks how they did, wants to prep for a follow-up, or asks what they missed. Also use on webinar, pitch and demo transcripts, and before a meeting to build a battle card on who is on the other side.

Run the user's semester from a machine-readable ledger built out of their syllabi. Use when they mention a class, an assignment, an exam, a deadline, a professor, a reading, a grade, attendance, or a syllabus; when they ask what is due, what to work on, or whether they can miss a class; when they hand over a syllabus PDF; and before helping with any graded work, because each course sets its own rules about what help is allowed.

Find the root cause of a bug instead of guessing at it. Use when the user reports something broken, pastes an error or a stack trace, says this used to work, asks why is this happening, why is this failing, or what is wrong with this. Also use to trace how data flows through a codebase, follow a request end to end, or find where a value is being lost or changed. Also fires on: stack trace, traceback, exception, crash, crashed, segfault, null pointer, undefined, regression, reproduce, bisect, ...

Research a topic, market, company or claim properly, with sources that can be checked. Use when the user asks you to research, look into, compare, or find out about something, when a decision needs evidence, or before writing anything aimed at an audience outside this machine. Not for a single fact lookup.

Record a product demo video by reading the product's own code and the craft of demo-making, not by guessing at either. Use when the user asks to record a demo, make a demo video, film the product, show the product working, or produce a launch clip. Also use when a demo already recorded is wrong, boring, or missed the point of the product. Needs the repo and a running instance.

Teaches an agent graph engineering, both halves: knowledge graphs (ontology design, entity/relation/event extraction, fusion, GraphRAG/memory serving) and task graphs (agent orchestration, parallel fan-out, verifier separation, the stop rule, human gates). Use when asked to build a knowledge graph, extract entities or relations from text, design an ontology, dedupe or merge entities, add graph memory or GraphRAG to an agent, orchestrate multi-agent workflows as a graph, or LEARN graph enginee...

Draw live interfaces on the screen over everything else, with no browser and no window. Use when the user asks to see something rather than be told it, says show me, put that on screen, pull up, or asks for a dashboard, chart, diagram, or status display. Also use to update or animate something already on screen, and to take it down when they are done.

Build any interface by loading a preset instead of re-deriving one: dashboards, tables, forms, detail views, settings, wizards, feeds, HUD panels, landing pages. Use when asked to build, show, visualize, lay out or mock up a screen, dashboard, chart, UI or app, when the user wants to see data rather than be told it, and before generating any UI from a data source. Also fires on: looks bad, looks generic, ai slop, slop, ugly, redesign, restyle, make it look, landing page, hero section, compone...

Build a kit, a repo somebody lives inside for weeks while an agent walks them through a process they have never done and will not do again for years. Use when the user faces a long multi-session process with a deadline and an evaluator (applications, appeals, accommodations, estate admin, a job search, a fundraise, a thesis), when they ask for help with something bureaucratic they have never done, when they want to hand a process to a friend, or when they say "make a kit". Also use to decide ...

Get to know someone without making them write an autobiography. Use in a first session, when Claude does not know enough about the person to be useful, when they offer files or a Drive export, when they say Claude keeps forgetting things, or any time an answer would be better with context you do not have.

The half of a week that is not code and not class: appointments, renewals, forms, commitments made in passing, the weekly review, and deciding what to drop when the week does not fit. Use when the user is planning a week, is overloaded, asks what they are forgetting, mentions an appointment or a deadline that is not academic, wants a review, or is deciding whether to take something else on.

Check a purchased, scraped or inherited contact list before anyone builds a sequence on it. Use when a lead list, investor list, prospect file, CSV export or enriched dataset arrives, when someone says they bought a list or paid for data, before writing outbound copy or scoring rules against a file, and when a campaign is bouncing or the personalization is coming out wrong. Also use before quoting how many contacts a file contains.

Click, type, drag, scroll, press keys, and drive UI on the user's Mac. Use when the user asks to click something, fill in a field, press a shortcut, select a menu item, or perform any action in a Mac app.

Drive specific Mac apps through AppleScript, JXA, Shortcuts, or their own data files - Mail, Messages, Notes, Safari, Calendar, Reminders, Finder, Music, Terminal. Use when the user asks to do something in a named Mac app, especially an Apple one.

Generate the morning operator brief - triage the user's email, texts, and calendar into what is urgent, what needs a reply, and a suggested order for the day. Use when the user asks for their brief, their morning rundown, what they need to handle today, what is urgent, or who is waiting on them. Runs on a schedule at 8am via launchd.

Control the user's Mac - click buttons, read the screen, drive apps, automate anything in the UI. Use when the user asks to click, open, close, read what is on screen, fill something in, navigate an app, or automate a task on their machine. Routes the request to the cheapest control layer that can do it, which is almost never a screenshot.

Diagnose a Mac automation that is failing, especially one failing silently. Use when a click does nothing, typing does not appear, the accessibility tree is empty, an AppleScript hangs or errors, coordinates land in the wrong place, or something that worked before stopped working.

Scan the user's texts, email, or calendar for things that need a response or an action, and turn them into follow-up tasks. Use when the user asks what they are forgetting, who is waiting on them, what they owe people, to catch up on messages, to follow up on something, or to turn their inbox into a to-do list.

Diagnose and fix macOS permission problems for automation - Accessibility, Screen Recording, Automation/Apple Events, Full Disk Access, TCC. Use when a Mac automation fails silently, when the user gets "not authorized" errors, when setting up Mac control for the first time, or when something worked before and stopped.

Run a multi-step task on the Mac as a formal plan that is type-checked, executed, verified, and logged - instead of improvising bash step by step. Use for any task with more than two or three steps, anything that ends in an irreversible action (send, pay, post, submit), or anything the user will want to debug afterward. This is Chewbacca's agent runtime.

Read what is on the user's Mac screen - the accessibility tree, window contents, UI elements, or a screenshot. Use when the user asks what is on their screen, what an app is showing, to find a button or field, or to check the state of something before acting on it.

Remember everything about the people in the user''s life, answer questions about their network, and notice who is slipping. Use when they mention a person by name, tell you something about someone, ask what you know about someone, ask who to reconnect with, mention a birthday or job change, or come back from a meeting or call. Use before drafting any message to a named person, so the draft is grounded in what is true about them. Use for last-contact questions: when did I last talk to them, ho...

Build a brief on a person or company before a call, pitch or intro. Use when the user names someone they are about to meet, has a meeting or call with them tomorrow or later today, asks who someone is, asks what to say to them, is preparing outreach, or wants to know whether a lead is worth the hours. Also use before drafting any message to a named person.

Find what has rotted in a project. Use when the user asks to clean this up, find dead code, remove unused imports, check my dependencies, whether anything is out of date or vulnerable, what is safe to delete, or asks for an audit of a project or repo. Also use when a repo feels messy, before handing a project to someone else, or when env vars have drifted.

Review code for real problems before it goes anywhere. Use when the user asks to review this, look over my code, check this for bugs, tell me if this is any good, find what is wrong with this, review my PR, or asks whether a change is safe to merge or ready for someone else to read. Also use before opening a pull request, after finishing a feature, or when a reviewer has called something sloppy.

Read from and write to the user''s personal context repo, the private markdown store holding who they are, what they are working on, who they work with, and how their systems are wired. Use when the user asks what you know about something, states a durable fact worth persisting (new role, project shipped or died, collaborator joined, preference, infrastructure change), says "remember this", asks for a context review, or when work just changed something context should reflect. Also use before ...

Install Chewbacca by talking, not by answering a terminal wizard. Use when the user wants to set up this kit, run setup.sh, finish a half-finished install, or add one tool that was skipped. Also use when they ask what the installer is about to change on their machine.

Get a change safely out the door. Use when the user says ship it, is this ready, push this, deploy this, put it live, run the tests, write tests for this, or asks whether something is safe to release. Also use before any deploy to production, when a build or deploy fails, and at the end of a work session to commit and push what was done.

Write a new skill, or fix one that is not being obeyed. Use when the user asks for a new skill, says a skill is being ignored or does the wrong thing, wants to turn a repeated task into something reusable, or asks to install a skill someone else wrote. Also use before writing any SKILL.md, because the authoring order matters more than the file.

Update a skill from how a real run actually went. Use right after a skill produces output the user corrects, rejects, or edits before using. Also use when the user says a skill keeps making the same mistake, or asks to make something remember a preference. This is the loop that makes skills improve instead of drifting.

Enforces this kit's standards for Next.js, React, Supabase and Vercel work. Use when writing or reviewing UI components, API routes, database schemas or queries, deployment config, styling, animations, accessibility, performance, state management or tests, or when running a pre-ship audit. Covers shadcn/ui patterns, Zod-validated App Router routes, RLS policies, Core Web Vitals budgets and WCAG contrast.

Build an interactive study guide, review sheet, practice quiz, walkthrough or flashcard set for a course topic, from lecture notes, slides, a transcript, a reading or a problem set. Produces one HTML file with quizzes, step-throughs and flashcards that records what was missed, so a later session picks up there. Use when asked to make or extend a guide, turn notes into practice, or review before an exam.

Turn course material into durable knowledge instead of a reread. Use when the user is studying for an exam or quiz, has lecture notes or slides to process, is assigned a reading, asks to be quizzed or tested, wants flashcards or a study plan, just got a graded exam back, or says they studied and still did badly.

Read, search, and remember the user's iMessage history. Use when they ask what someone said, what they talked about, when they last spoke to someone, what they missed, or to catch up on a thread. Also use after any conversation that mentions a text, so what mattered in it gets written down before it scrolls away. Also fires on: my texts, read my texts, text messages, iMessage, imessages, texted, texting, chat history, message thread, DMs, what did they text me.

Propose things this kit could do here that the user has not thought to ask for. Use when the user asks what you can do, what they should automate, what they are missing, or says they do not know where to start. Also use on arriving in an unfamiliar repo, and when the user is stuck choosing what to work on.

Decide what Chewbacca is allowed to read, then get it in. Use when somebody has just installed this and does not know where to start, asks what it can see or read, asks how to get their life into it, asks which permissions it needs or why macOS is prompting, wants to import messages, contacts, LinkedIn, email, calendar, notes or files, or asks what to do with a Google Drive or Takeout export. Also use when a question could not be answered because the data was never imported, and when deciding...