Field notes

The ClaudeKit blog

Measured token costs, kit teardowns, and working setups for Claude Code. We publish the numbers behind every claim.

Claude Code Kits vs Free Prompt Packs: What $14.99/mo Actually Buys

Free Claude Code prompt packs vs paid kits — an honest comparison of what you get, what breaks, and when free is genuinely the right call.

June 6, 202611 min read

Read the post

Guides

Getting Started With ClaudeKit: Install a Kit in 5 Minutes

From checkout to first slash command in under 5 minutes — native plugin install or the ck CLI, step by step. Covers all 5 kits and 101 commands.

June 6, 202610 min read

Data

Measuring Claude Code Context Token Costs: The Full Breakdown (2026)

Every skill, command, and agent you install costs context tokens on every session. How ClaudeKit measures 82,197 tokens across 5 kits and how to audit your own setup.

June 6, 202612 min read

Guides

What Are Claude Code Subagents? (And Why Specialists Beat Generalists)

Claude Code subagents are specialized AI agents with their own context, tools, and instructions. Here's how they work and when to use them in 2026.

June 6, 202613 min read

Guides

Six Months Building ClaudeKit in Public: What We Shipped, What We Killed, and the Numbers Behind v2

A June 2026 build-in-public retrospective: 5 kits, 101 commands, 82,197 measured tokens, and the architecture decisions that got us here.

June 4, 202611 min read

Guides

From One YouTube Video to 12 Assets: The Repurposing Matrix

How MarketingKit's /mkt repurpose command turns one long-form video into 12 platform-native assets in your cloned voice — Shorts, threads, newsletter, LinkedIn, and more.

June 3, 202611 min read

Data

The Real Cost of Free Skills: Token Waste in Popular Repos, Measured

We measured the context cost of free SKILL.md files using chars/4. The pattern: always-loaded bloat and missing progressive disclosure. Here's the math.

June 1, 202613 min read

Comparisons

AEO vs SEO: Optimizing for AI Answers Without Abandoning Google (2026)

AEO optimizes to be cited in AI answers; SEO optimizes for clicks from search. They overlap on structure, diverge on the goal. Where, and how to measure each.

May 28, 202612 min read

Guides

Claude Code for Ecommerce: The Missing Operations Manual

EcomKit gives Claude Code 20 commands, 3 skills, and 2 agents covering the full store spine — catalog to PDP to lifecycle to ads to marketplaces. 16,464 measured tokens.

May 25, 202611 min read

Guides

Batch Video Generation: 200 Personalized Remotion Renders from One CSV

VideoKit binds a CSV to a parametrized Remotion composition and fans out 200 renders locally or on Lambda. Pipeline, naming conventions, cost breakdown, and provenance policy.

May 21, 202613 min read

Guides

How Much Context Does Your ~/.claude Folder Eat? Run This Audit

A copy-paste bash audit estimates how many tokens your ~/.claude skills load. Interpret the numbers, prune the bloat, then automate with ck tokens/doctor.

May 18, 202612 min read

Guides

How to Optimize Amazon Listings with Claude Code in 2026

EcomKit's /ecom amazon command writes byte-aware titles, five benefit-led bullets, A+ modules, and backend keywords, then runs a compliance flag pass before you submit.

May 14, 202612 min read

Comparisons

Claude Code Plugins vs Kits: What the Marketplace Misses

Claude Code plugins give you broad, free, composable parts. Curated vertical kits trade breadth for orchestrated workflows, a quality floor, and measured token cost.

May 7, 202612 min read

Guides

The llms.txt Playbook: How to Get Your Product Cited by ChatGPT and Perplexity

llms.txt gives AI assistants a clean map of your site. Learn what it is, who reads it, how to generate and maintain it, and how to measure AI citation share.

May 4, 202614 min read

Guides

YouTube Scripts with Claude Code: Hook, Retention, CTA Architecture

Write a full YouTube script in Claude Code with VideoKit: a 30-second hook, retention curve with re-hooks every 40-60s, and CTA placement, in your voice.

April 30, 202613 min read

Guides

Klaviyo Flows Written by Claude Code: Welcome, Abandon, Winback Templates (2026)

Generate a full Klaviyo lifecycle program in Claude Code with EcomKit: welcome, abandoned-cart, browse-abandon, post-purchase, and winback flows with real copy.

April 27, 202611 min read

Guides

Haiku, Sonnet, or Opus? Model Tiering Your Claude Code Workflows for 10x Cost Savings

Match the model to the task: Haiku for mechanical work, Sonnet for craft, Opus for strategy. The framework, cost math per 1,000-task month, and real failure cases.

April 23, 202612 min read

Comparisons

wshobson/agents Review: 192 Free Agents, Honestly Assessed (2026)

wshobson/agents has 192 agents, 84 plugins, four model tiers, and multi-harness export. An honest architecture review and when to pay for a vertical kit instead.

April 20, 202613 min read

Guides

Claude Code Slash Commands: Orchestration Patterns That Actually Work (2026)

Five orchestration patterns behind ClaudeKit's 101 slash commands: namespaced verbs, bootstrap-context-first, parallel fan-out, evidence contracts, and diagnostic vs. deliverable split.

April 13, 202613 min read

Guides

Why We Killed the Reviewer Gate (and What We Use Instead)

ClaudeKit v2 dropped blocking reviewer/quality-gate agents. Commands now end with EVIDENCE — a diff, report, or verified file — not a gated loop. Here's why.

April 9, 202612 min read

Guides

How We Watermark Text Products (and Why Kit Resellers Hate It)

We watermark every ClaudeKit download three ways: per-buyer LICENSE_HOLDER.md, licensed-to frontmatter, and HMAC-signed URLs. The honest playbook.

April 6, 202614 min read

Guides

Programmatic SEO with Claude Code: Template + Dataset Design That Ranks

Programmatic SEO that ranks needs a template married to a unique dataset, with thin-content guards. Design, tools, and a worked example using SEOKit v2.

March 30, 202614 min read

Guides

Remotion + Claude Code: Generate Product Demo Videos from a Prompt (2026)

How to scaffold a Remotion project, build a composition from a brief, add captions, and render — using VideoKit's 17 commands and 5 skills.

March 26, 202613 min read

Comparisons

The 10 Best Free Claude Code Skill Repos (Honestly Ranked, 2026)

Best free Claude Code skill and agent repos ranked: anthropics/skills, wshobson/agents, davila7 templates — honest strengths, tradeoffs, and token costs.

March 23, 202612 min read

Guides

How Claude Code Runs a Full SEO Team: The v2 Architecture Explained

SEOKit uses 19 commands, 4 skills, and 2 read-only specialist agents to produce ranked, falsifiable SEO backlogs — no orchestrator, no reviewer gate.

March 19, 202611 min read

Guides

How to Stop Claude Code Context Spirals: A Token Budget Playbook (2026)

Claude Code burns context on MCP schemas, global CLAUDE.md, and idle skills before your task even starts. Audit, measure, prune, and tier models to fix it.

March 16, 202612 min read

Comparisons

ClaudeKit Review 2026: Are the 5 Vertical Kits Worth It?

Honest ClaudeKit review: 5 kits, 101 commands, 82,197 measured tokens. Pricing, install steps, what each kit does, and who should skip it.

March 12, 202611 min read

Guides

Claude Code Agents vs Skills vs Slash Commands: When to Use Which

Skills are model-triggered capabilities, agents are read-only specialists with their own context, slash commands are user-triggered workflows. Exactly when to use each.

March 9, 202613 min read

Guides

The Complete Guide to Claude Code Skills in 2026

What Claude Code skills are, how they differ from slash commands and agents, and how ClaudeKit's 5 kits (82,197 tokens measured) use them to ship real work.

March 5, 202614 min read

Data

What Do Claude Code Skills Actually Cost in Tokens? We Measured the Full ClaudeKit Catalog

We measured all 82,197 tokens across ClaudeKit's 5 v2 kits. Typical skill: 600–1,500 tokens. Here's the full breakdown by kit, band, and what to cut first.

March 2, 202612 min read