Jeff-Kazzee

Design slide carousels and slide decks for social: Instagram carousels, LinkedIn document/PDF carousels, and short explainer slide decks. Use whenever the user asks to "make a carousel", "instagram carousel", "linkedin carousel", "carousel about X", "slides for a post", "swipe post", "turn this into slides", or a "slide deck" for social. Part of the BANGERS suite; applies one-idea-per-slide clarity and strong-first-slide packaging. Reach for this any time the deliverable is a multi-slide swipeable/scrollable visual post.

Produce editing and recording assets for any editor — DaVinci Resolve, Adobe Premiere, CapCut, Descript — plus OBS for recording: cut sheets / edit decision lists, burned-in caption scripts and SRT subtitle files, marker/chapter CSVs, B-roll and text-overlay cue sheets, vertical reframe plans, and OBS scene/recording setups. Use whenever the user asks to "edit this video", "davinci", "resolve", "premiere", "capcut", "descript", "obs setup", "how should I cut this", "cut sheet", "edit decision list", "captions for my video", "make an SRT", "add markers", "reframe to vertical", or wants a recording/editing plan. Part of the BANGERS suite. This skill outputs instructions and files the user feeds into their editor and OBS (it does not drive an NLE directly).

Write scroll-stopping hooks, video titles, and thumbnail concepts. Use whenever the user asks to "write a hook", "give me a better opening line", "title ideas", "come up with a title", "thumbnail idea", "thumbnail concept", "how should I open this", "make this more clickable", "my video isn't getting views", "help me with packaging", or is polishing the first 3 seconds / first line / title / thumbnail of any piece. Part of the BANGERS suite. Packaging (hook + title + thumbnail) often decides whether content is ever seen, so reach for this skill aggressively any time the opening or packaging of a piece is in play.

Generate content ideas, angles, hooks, and a posting plan for a creator's social channels. Use this whenever the user asks "what should I post", "give me content ideas", "I need video ideas", "help me brainstorm content", "what should I make this week", "content calendar", "I'm out of ideas", "ride this trend", "turn this news into content", or wants a batch of angles on a topic or niche. Also use proactively at the start of any content session to pick WHAT to make before HOW to make it. Part of the BANGERS social-media suite; applies the idea-selection tactics of Fireship, Theo, ThePrimeagen and Matt Pocock, adaptable to any creator's niche.

Write long-form written posts: LinkedIn posts, Substack newsletter issues, and Facebook Group posts. Use whenever the user asks to "write a LinkedIn post", "write my newsletter", "substack post", "newsletter issue", "facebook group post", "a post for my community", "turn this into a LinkedIn post", or wants professional/newsletter/community written content with a hook, body, and CTA. Part of the BANGERS suite; applies the teaching clarity and story-led hooks of the reference creators, adaptable to any creator's niche. Reach for this for any written post destined for LinkedIn, Substack, or a Facebook Group.

Turn ONE piece of content into a full cross-platform package — the "film once, cut everywhere" waterfall. Use whenever the user asks to "repurpose this", "turn this video into posts", "post this everywhere", "make content from this", "one idea for all platforms", "content from my podcast/stream/video", "atomize this", "give me everything for this topic", or hands over a source (video, transcript, script, article, idea) and wants it adapted across YouTube, Shorts, TikTok, Instagram, X, LinkedIn, Substack, Bluesky, and Facebook Groups. Part of the BANGERS suite; the orchestrator that routes work to the other banger skills.

Write full long-form video scripts for YouTube (and long-form talking-head/tutorial/reaction content). Use whenever the user asks to "write a YouTube script", "script for my video", "long-form video about X", "write a video explaining Y", "tutorial script", "reaction video script", "10-minute video on Z", or needs a complete narrated script with a retention structure, hook, chapters, and B-roll/visual cues. Part of the BANGERS suite; applies Fireship's density, Matt Pocock's teach-one-thing clarity, and Theo/ThePrimeagen's reaction-and-verdict energy, adaptable to any creator's niche. Reach for this any time the deliverable is a spoken script longer than ~90 seconds.

Write short-form vertical video scripts for YouTube Shorts, TikTok, and Instagram Reels. Use whenever the user asks to "write a short", "TikTok script", "Reel script", "make a Short about X", "60-second video", "vertical video script", "hook for a short", or wants a sub-3-minute punchy vertical script with captions and on-screen text. Part of the BANGERS suite; applies Fireship-style density and the first-3-seconds discipline all four reference creators share, adaptable to any creator's niche. Reach for this any time the deliverable is a short vertical video.

Build or refine a truthful, memorable personal brand positioning system — positioning statements, bios, content pillars, proof-asset plan, anti-positioning. Use when the user says "work on my positioning", "personal brand", "rewrite my bios", "what should I be known for", or "is my positioning still right".

Improve the Growth Engine itself — review which advice worked and which didn't, interview the user about friction, and update the playbook that every other skill obeys. Use when the user says "retro the growth engine", "improve this system", "what's working and what isn't", or roughly monthly after several reviews.

Run an accountability review of the user's growth system — score follow-through on past commitments, kill stale bets, check the proof-asset pipeline and learning roadmap, flag drift from goals. Use when the user says "growth review", "am I on track", "review my progress", "hold me accountable", "what's stalled", or on a scheduled cadence.

Set up or update the Growth Engine memory folder and builder profile. Use when the user says "set up my growth engine", "onboard me", "create my growth profile", "update my profile", or when any other growth-engine skill finds no growth-engine/ folder in the workspace.

Deep-research a niche and turn it into a living markdown knowledge base (market map, personas, pain points, tools, datasets, project ideas, content angles). Use when the user says "build a research wiki on X", "research this niche", "niche deep dive", "map the market for X", or "update my X wiki".

Discover, verify, rank, and package emerging opportunities for software, AI workflows, tools, directories, automations, or small online businesses. Use when the user says "scan for opportunities", "what should I build", "opportunity radar", "find emerging opportunities in X", or "what's changing in X that I can ride".

Use after the spec is clear and before implementation, when a coding task involves real architecture or design trade-offs - data model, API shape, build-vs-buy, a new module, choosing a library, or how the agent's context and guardrails are set up. Make sure to use this whenever you ask "how should we structure this", "what's the architecture", "which library", or faces any non-trivial structural decision, even if he doesn't say "design". Second step of the SDLC loop; skip it for trivial changes.

Use during active implementation of a planned, tested feature - once a spec and failing tests exist and it's time to write code. Drives the agent to implement against the spec + tests, reserves your attention for the hard 20% (ambiguity, edge cases, correctness), and reviews code as it appears instead of accepting it blind. Make sure to use this whenever you are coding a feature that already has a spec/tests, or says "now build it", "make the tests pass", "implement the slice". Fourth step of the SDLC loop.

Use on a weekly cadence, or after a rough project, to turn logged failures into durable improvements. Reads the agent-run log, clusters recurring agent failures, encodes each into an AGENTS.md rule or checklist, and updates the KPI tracker so the trend is visible. This is the compounding self-improvement step that makes you measurably better at directing agents over time. Make sure to use this whenever you say "weekly review", "retro", "what did we learn", "update my rules", or "let's improve the loop". Final step of the SDLC loop - it feeds the next cycle.

Use at the START of any coding task, feature, or project - before design or code. Turns a rough idea or request into a clear, testable spec that clarifies real intent, surfaces edge cases, and deliberately hunts for the simplest path to the same outcome (fewest steps, least complexity). Make sure to use this whenever you say "let's build", "I want to add", "new feature/project", "plan this out", or hands over a vague coding request - even if he doesn't say the word "plan". First step of the SDLC loop.

Grade agent output before it ships, and log it. Use when agent-produced code is about to ship - a PR, a merge, a deploy, or "is this ready". Runs deterministic gates first, then grades the change against a short rubric (spec adherence, error handling, approach, reviewability), and logs the run so the self-improvement loop has data. This is the ONLY place an LLM judge is allowed, and only for code review. Make sure to use this whenever you say "review this", "before I ship/merge", "is this ready", or hands over a diff/PR. Fifth step of the SDLC loop.

Use BEFORE generating implementation code for any feature with verifiable behavior. Writes the tests (and evals, if output quality is judgment-based) FIRST, as the contract that defines "correct," so the agent implements against a real target instead of a vibe check. Make sure to use this whenever you say "implement", "write the code for", "build the function/endpoint/feature" - intercept and do tests first - and also on "write tests" or "add coverage". Third step of the SDLC loop and the single highest-leverage habit.

Help an overwhelmed or uncertain beginner decide how to start using AI for life, work, school, a small business, or creative projects. Use for requests such as "I want to get into AI," "I do not know where to start," "which AI tools should I learn first," or "make me a beginner AI plan." Produce a tiny tool stack, a seven-day plan, one useful workflow, and one practice project. Do not use for advanced ML engineering, a single broken prompt, enterprise AI strategy, or a job-search-only request.

Generate and rank small, realistic AI project ideas for a beginner, non-coder, self-taught developer, creator, job seeker, freelancer, worker, or small-business owner. Use for "what should I build with AI," "give me a weekend project," "I need a portfolio project," "small AI app ideas," or "something useful I can automate." Match ideas to interests, skills, tools, time, budget, and audience; recommend one MVP and a seven-day build plan. Do not use for unconstrained generic brainstorming, full startup market research, enterprise strategy, or career planning when a project is not the main need.

Help a user turn AI learning into honest career progress, portfolio evidence, workplace value, job-search assets, or a realistic freelance offer. Use for "how do I get a job in AI," "what AI skills belong on my resume," "how do I prove I can use AI," "build my AI portfolio," or "how can I freelance with AI." Audit current proof, skill gaps, claim strength, and target opportunities; then create a 30-day evidence-building plan. Do not fabricate credentials, guarantee employment or salary, apply or contact people without approval, or replace legal, benefits, immigration, or financial advice.

Create a realistic AI learning path tailored to a user's target outcome, starting level, schedule, learning style, tools, and access needs. Use for "teach me AI," "make me a learning plan," "help me learn prompting or agents," "I want AI for coding or my job," or "I only have 30 minutes a day." Build milestones, practice, proof artifacts, and review checkpoints. Do not use for one-off questions, current job searches, project brainstorming alone, or broad career coaching when learning is not the main issue.

Diagnose and repair a prompt that produces vague, generic, incorrect, badly formatted, or model-mismatched results. Use for "fix this prompt," "why did the AI misunderstand me," "rewrite this for Claude/GPT/Gemini," "the model ignores my format," or "my prompt is not working." Identify whether the real problem is instructions, context, examples, output structure, model choice, missing data, or missing tools; then rewrite and test it. Do not use for general writing feedback, one-off factual questions, full automation architecture, or broad career coaching.

Turn a large software project or feature set into epic-scoped developer loops with testable task contracts, local or cloud CI/CD gates, review/fix cycles, PR babysitting, and human approval only at real gates. Use when planning how agents should decompose, build, verify, review, document, and hand off a large project without bothering the human after every passing task.

Use when creating, revising, or reviewing Obscura branding assets, SVG logo/icon/favicon work, website images, screenshots, social previews, or Loom/hero motion video assets.

Use when updating Obscura code, docs, GitHub issues, releases, or roadmap state and the work needs to be logged both locally and on GitHub. Keeps DEVLOG, CODEMAP, release notes, issue comments, and prod/dev branch state consistent.

Reduce work in a Vivary workspace to first principles, preserve context-window headroom, and communicate the shortest clear path in light caveman-style language. Use when the user says "Grug", "Grug mode", "caveman", "keep it simple", "first principles", "preserve headroom", "stop overthinking", or asks for an unusually terse plan, explanation, implementation slice, review, or status update.

Bootstrap, maintain, and continuously improve a Vivary agent workspace so each session adds momentum the next inherits. Lays down the workspace's shared brain (SOUL/USER/AGENTS/MEMORY/STATE) from strato's templates, runs periodic heartbeats to surface stale folders and skill candidates, captures recurring bugs into a self-healing playbook, and extracts a skill once a workflow repeats a third time. Use when starting a new workspace, when something keeps biting, when a workflow has fired three times, or when the user says "bootstrap the workspace", "set up the brain", "run a heartbeat", "audit the workspace", "tend the workspace", "this happened again", or "strato".

Background pattern-finding for a Zo Computer workspace, modeled on Anthropic's Dreaming feature for Claude Managed Agents (released 2026-05-06). Runs between sessions, reviews recent activity across projects plus Files, Skills, Personas, Rules, Automations, and hosted Sites, and surfaces cross-project bleed, drifted Rules, unused Skills, and other patterns the agent can't see in any single session. Asks the user before any destructive action. Triggers on "run dreaming", "let zo dream", "dream over my workspace", "consolidate the workspace", "what's drifting", or on a scheduled Automation.

Bootstrap a Throughline WS with interview gates, PRIV files, VS, source trail, checks, and update cadence.

Use this skill when the user wants career coaching, job-search help, resume/profile improvement, interview prep, promotion planning, performance tracking, skill-gap analysis, career-change planning, freelance positioning, or recurring career progress reviews. Do not use it for generic productivity advice unless the advice is directly tied to career outcomes.

Typecheck and fix YAML frontmatter across a Markdown knowledge base / vault (Obsidian, docs, notes). Use when the user wants to validate frontmatter, enforce a metadata schema, find Markdown files with missing or malformed frontmatter, audit a vault's structure, add a new document type, or scaffold frontmatter onto files that lack it. Triggers: "check my frontmatter", "validate the vault", "enforce a schema on my notes", "what types of docs do I have", "fix missing frontmatter", "lint my Markdown metadata".

Bootstrap, maintain, and continuously improve a coding-agent workspace so each session adds momentum the next session inherits. Sets up the shared brain (SOUL/USER/AGENTS/MEMORY plus agents inbox + group chat), runs periodic heartbeats to surface stale folders and skill candidates, captures lessons into a bug-risk playbook (self-healing), and proactively packages repeated work patterns into skills, hooks, commands, or workflow plugins. Use when starting a new workspace, when something keeps biting and needs to land in a playbook, when a workflow has repeated three times, or when the user says "spin up the flywheel", "run flywheel", "audit the workspace", "tend the workspace", "set up the workspace", or "run a heartbeat".

Reads subscription receipts in the user's Gmail, on the user's own Zo. Prices the waste, stamps KILL, KEEP, or TRIM on every bill, and prints a receipt. Nothing personal leaves this machine.