agt-skill-pack

Turn a content creator's AI stack into a factory that sounds like THEM. One voice spine (CONTENT.md) governs every format — FB/blog posts, TikTok carousels, YouTube Shorts, covers — so the output stops smelling like AI and comes out consistent every run. Use when a creator (not a dev) wants to set up Claude Code / an AI agent for content production across channels, or fix output that reads generic, breaks Thai fonts, or looks different every gen. Triggers: "setup AI for content", "brandbook for AI", "AI writes generic", "content creator claude md", "carousel/shorts pipeline", "cover generator".

Keep an AI character's face and voice identical across every clip. Lock the identity ONCE (train a face, lock a voice), then reuse it — never re-generate the face. Use when a recurring AI character drifts between clips, looks like a different person each render, or the brand face is never consistent.

Split AI content into two layers — AI generates raw material (backgrounds, characters, clips), code assembles the deliverable (text, layout, render). Use when AI-generated carousels/cards come out inconsistent, Thai fonts break, or every re-gen is a gamble.

Pump the brakes — clarify before acting. Use when a request is exploratory, mid-decision, or underspecified and the AI is about to edit, commit, hand off, deploy, or run a workflow on a partial picture. Fires on /hold-up or "just talk", "don't action this", "let's think first" — or auto-fires when the AI catches discuss-mode cues right as it was about to take a non-trivial action.

Game-theory review loop for AI drafts: tit-for-tat trust tracking across review rounds. Review depth adapts to the author's track record of addressing feedback, so clean drafts earn lighter reviews and ignored flags trigger deep audits. Use for recurring content or code review loops.

Personal commitments tracker with escalating confrontation. Use when the user wants accountability on recurring commitments, training, habits, or deadlines they keep deferring. Reviews the track record and escalates tone based on the avoidance pattern, not the excuses.

Run multiple AI coding agents in parallel without collisions. Use when setting up parallel-agent workflows: defines file boundaries, a written contract between agents, and conflict rules so agents stop overwriting each other's work.

Give your AI long-term and short-term memory: keyword-searchable session history (FTS5 index) plus a hot cache of current project state loaded at session start. Use when AI should remember past sessions and know project state without re-explaining.

Detect whether a UI looks AI-generated. Use during design review to run the fingerprint test: identifies the default-theme, gradient, and layout tells that make interfaces read as AI slop, and how to fix each one.

Set up a CLAUDE.md that stops AI guessing, claiming done without verifying, expanding scope past the ask, and forgetting between sessions. Use when configuring Claude Code rules for a repo or workspace.

Detect when past decisions have gone stale because their inputs changed. Use to review a decision journal: flags choices whose assumptions the world has moved past, and forces a revisit-or-reaffirm call.

Pattern-recognition design review: judge whether an interface works before any analysis. Use for UI and design reviews, visual hierarchy checks, and 'does this look right' calls.

Cross-model adversarial critique: use a different AI model to review anything your primary AI produced — posts, specs, code, research, decisions. Same-model self-review falls into local minima; a different model catches different blind spots.

Bootstrap Claude Code for React/Next.js repos: project context, conventions, and guardrails so it stops guessing and builds correctly from the first prompt. Use when setting up Claude Code in a frontend codebase.

Turn dropped URLs into summarized, categorized entries in a persistent reading queue with priority scoring. Use whenever links should be captured, summarized, classified, and queued instead of rotting in bookmarks.

Set up a reliable daily live-data brief: deterministic fetch scripts feed AI clean numbers instead of letting it browse half-loaded pages while writing. Use for morning market or metrics summaries.

The simpler-thing test: before a fix ships, prove the smaller version doesn't work. Use when code feels overbuilt, when complexity needs challenging, or when asking 'is this necessary'.

Structured feature evaluation before enthusiasm or inertia makes the decision. Use for 'should we build this' calls: what to build, what to cut, and the test that separates the two.

Scan GitHub trending daily and report which repos contain ideas worth stealing for your project — patterns, not stars. Use for repo scouting and idea extraction.

Scout X/Twitter for what's already been said on a topic before writing: what performed, what's saturated, what angle is missing. Use for pre-writing research and content gap analysis.

Pre-publish content intelligence skill for marketing teams and content creators. Before drafting any post, scout what's already performing on X/Twitter and across platforms. Uses Twitter CLI for real-time data and Manus for cross-platform deep research. Returns a structured report: what's trending, what's saturated, where the gaps are, and what angle to take.

Workflow Architect. Socratic interview → workflow map → AI setup recommendations → vertical prompt scouting. Works on any AI chat (Claude, ChatGPT, Gemini) and Claude Code (install skill).