melek-skills

Methodology for designing, refactoring, and evaluating AI agents — when to use a tool vs a skill vs a subagent, how to keep system prompts lean, and how to improve agents through eval-driven hill climbing. Use this skill whenever the user is building a new agent, refactoring or debugging an existing one, complains that an agent "got worse" or "is inconsistent", asks about system prompt length, tool design, multi-agent/subagent architecture, or wants to create evals for an LLM app — even if they don't use the word "agent" (e.g., "my Claude workflow keeps messing up", "should I add another tool?", "how do I test my prompt changes?").

Run an adversarial review→fix loop on a GitHub PR using BMAD adversarial agents until the reviewers find nothing actionable, then stop and report. Never auto-merges. Use after opening a PR, or when the user says "adversarially review", "hammer this PR", "review until clean", or /adversarial-pr-review.

Generate and send invoices through the Invoice Ninja API. Triggers on invoice creation, billing, generating a quote/devis.

Write professional emails with proper context gathering before drafting. Triggers on email writing, reply drafting, cold outreach.

Resolve git merge conflicts in an Obsidian vault that's synced via git (e.g. across multiple devices, or with iCloud/Drive on top). Defaults to keeping the most recent block in daily notes.

Brand guidelines, voice, and design system for Clip2Earn — EMEA creator program infrastructure (sub-brand of Bakchich). Use when creating ANY content or UI for Clip2Earn — pitch decks, presentations, blog posts, ad copy, social captions, video scripts, creative briefs, landing pages, email campaigns, web app screens, or component code. Covers both audiences (creators + brands/SaaS founders), two distinct voice calibrations, the full HeroUI-based color system, Chakra Petch/Inter/Fira Code typography, shadows, radii, and iconography. Trigger this skill whenever the user mentions Clip2Earn, clip2earn.money, or asks to write content / build UI for an underground UGC marketplace OR a B2B creator-program infrastructure pitch.

Use when setting up or optimizing a repo for Claude Code — slim CLAUDE.md, scoped rules, RTK install, model routing, token reduction

Use when auditing or optimizing an ecommerce store for COD (Cash on Delivery) markets — MENA, North Africa, Pakistan, SE Asia. Covers checkout UX, trust signals, RTO reduction, COD-to-prepaid conversion, mobile-first design, and RTL architecture.

Use when publishing Instagram carousels from HTML files — covers export to PNG at 1080x1350px via Playwright and upload via Upload-Post API. Includes contrast check checklist before export.

Use when building a complete brand identity for a new or existing brand — brand strategy, voice document, moodboard, logo concepts, color palette, and full AI-ready brand kit. Triggers on "build my brand", "create brand identity", "brand guidelines", "brand kit", or when a project needs visual + copy foundations from scratch.

Use when building, auditing, or reviewing a startup landing page. Covers hero copywriting formulas, top-to-bottom section order, design tokens, trust signal hierarchy by growth stage, CTA playbook, tool recommendations, A/B testing, and analytics. Trigger on "landing page", "homepage", "conversion rate", "hero section", "above the fold", "CTA copy", or /landing-page-startup. Based on 18-source deep research (NotebookLM 082aa882, April 2026).

Use when planning, auditing, or optimizing Meta (Facebook/Instagram) ad campaigns for ecommerce or COD (cash on delivery) markets. Covers campaign structure, creative strategy, Advantage+, COD lead qualification, RTO reduction, and key metrics benchmarks.

Use when building self-improving autonomous systems, optimization loops, or applying experiment-driven iteration to any domain. Triggers on "auto-improve", "self-optimizing", "experiment loop", "karpathy loop", autonomous optimization, or when a workflow could benefit from measure-analyze-rewrite cycles.

Use when planning cold outreach campaigns, LinkedIn prospecting sequences, multichannel outreach architecture, or signal-based targeting. Covers LinkedIn + email sequencing, profile optimization, cadence design, deliverability, and campaign benchmarks. Also use when reviewing or optimizing existing outreach campaigns.

Multi-source deep research combining YouTube, web search, and article URLs into NotebookLM as a free RAG backend. Use when user wants to research a topic in depth, find up-to-date information, generate analysis, infographics, podcasts, quizzes, or any NotebookLM artifact. Triggers on "research X", "deep dive into X", "find everything about X", or /deep-research.

Run a lightweight GTD-style task system inside an Obsidian vault. Use when triaging inbox items, cleaning outdated task systems, clarifying tasks into daily notes/projects/waiting-for/watchlist, or maintaining an execution-first Obsidian workflow for a founder/operator. Especially useful when migrating from a heavier Notion or GTD setup into a simpler Obsidian flow.

Generate competitive battle cards from meeting notes — per-competitor positioning, advantages, objection scripts, and win/loss tracking. Optionally enriched with web research.

Analyze Discord support messages to find bugs, feature requests, and UX gaps. Clusters by topic, deduplicates against Notion tickets, creates new tickets for gaps.

Identify documentation gaps from prospect questions in meeting notes. Clusters recurring questions, ranks by frequency, cross-references with existing docs, and generates article briefs. Outputs markdown, Notion pages, or GitHub issues.

Generate a personalized sales deck for a prospect using real quotes and data extracted from meeting notes. Produces an 8-slide markdown deck with verbatims, not generic copy.

Orchestrator that runs all sales-agents skills in sequence — weekly deal review, battle cards, help center gaps — and compiles a unified weekly pipeline report. Supports full or partial runs.

Parse meeting notes to generate a weekly deal pipeline review with prospect data, budgets, objections, and next steps. Works with Gemini Meet notes, Claap transcripts, or raw text.

Search YouTube directly from Claude Code. Returns structured results with titles, channels, view counts, duration, and links. Use when user asks to search YouTube, find videos, or uses /yt-search.