Free · Try demo
Text Summarizer
Ollama swap included
NEW · 31-page Skills Report 2026 (free PDF) →
Skip the AI wrapper build. Own the working code.
Every wrapper you would otherwise pay $20-50 monthly for as SaaS. Working code, tested Ollama swap included, honest quality notes on where the paid API path still wins. 8 free apps to start, $99 bundle for all 15. One-time purchase, lifetime updates, own it forever.
Need a done-for-you agent build? Prompt Audit $499, Custom Build from $5K →
Need a workflow for your next proposal, meeting or draft? Claude Work Kit · $20 · Five tasks with files and checks →
58 recorded purchases across the site
RazorpayVISAMCAMEX
Instant deliveryThe Ollama swap
4 lines. Zero API bill. Runs on any laptop.
- import Anthropic from '@anthropic-ai/sdk';
- const client = new Anthropic();
+ import { Ollama } from 'ollama';
+ const client = new Ollama();
// Same prompts. Same shape.
// 80-85% of paid quality.
Monthly bill−100%on the workloads it handles
Works with•Claude Sonnet 4.6•Opus 4.7•Haiku 4.5•Claude Code 2.x
AI Wrappers · 8 free + 15 paid all live · $99 all-access bundle
Try 8 free apps in your browser. Own all 15 for $99.
Every wrapper you would rent as a $20-50/mo SaaS. Working code, Ollama swap included, honest quality notes. Skip the build, own the code.
8 free · live in-browser demos
Try them free first
All 15 paid live · $20 each or $99 all-access bundle
Or lock in the bundle for all 15
$20 · Ready to buy
PDF Q&A Bot
Upload PDF, ask questions with citations. FastAPI + Claude Haiku 4.5. Ollama swap included.
$20 · Ready to buy
Cold Email Personalizer
Paste LinkedIn profile, get 3 tone-calibrated cold email variants. Express + Claude + Groq swap.
$20 · Ready to buy
SEO Article Generator
Two-pass outline-then-fill for 1,000-word SEO articles with FAQ. Sonnet 4.6 quality, Groq swap.
Get the $99 all-access bundle → and lock in every wrapper we ship (12 more on the roadmap, all included).
For teams and founders
Building AI agents for your team? Done-for-you engagements from $499.
Prompt Audit at $499: send transcripts, get a written action plan in 5 business days. Custom Agent Build from $5,000: I design and ship the agent end-to-end for a specific workflow in your stack. Async only, fixed price before work starts, you own everything.
Claude Work Kit · $20 once / ₹1,199 in India
The expensive part of AI is the work you have to redo.
Five complete workflows for proposals, meeting notes, editing, decision memos and bug tickets. Each includes editable inputs, staged prompts, an example output and checks for common mistakes.
Prompts are available free. The kit gives you the prepared task materials and review steps together, plus the existing cheat-sheet reference.
Featured prompt packs · $45 to $49 each
Own the prompt structure that vertical SaaS charges you monthly for.
Each pack ships 50 tested Claude prompts for one profession, organized by the 4 roles every team has. One-time payment, lifetime updates, installable as Claude Code skills.
Basis, ABAP, Functional, Architect
SAP Pack
50 Claude Code prompts built for the four SAP roles that actually touch the system every day.
Own the prompts that SAP Build AI / specialized SAP AI consultancies charges $300-1,500/mo to rent.
Solo/GP, Litigation, Transactional, Practice Mgmt
Legal Pack
50 Claude Code prompts for the four legal roles that handle the actual billable work: solo / GP, litigation, transactional, and practice management.
Own the prompts that Harvey AI, Spellbook, CoCounsel, Paxton charges $249-1,200/mo per seat to rent.
.cursorrules, Slash Commands, Sub-Agents, Dev Prompts
Cursor + Claude Pack
50 power-user assets for shipping faster with Cursor and Claude Code: 15 .cursorrules files, 13 slash commands, 10 sub-agents, 12 dev prompts — all production-tested.
Own the prompts that Cursor Pro + Claude Code Pro + scattered free repos charges $40+/mo subscriptions to rent.
Full catalog
Every product on one page.
Paid wrappers, more packs, Sales Agent Pack, Skills Library, plugin, and services. All one-time, all lifetime updates.
Why this site exists
Tested 160+ popular Claude codes for three months. Most only change how it sounds.
Every “Claude pro tip” thread on the internet is someone's hunch. I got tired of arguing about it, so I built a rig, fresh contexts per trial, a fixed battery of 24 tasks across writing, coding, and analysis, outputs rated blind by three reviewers on decisiveness, accuracy, and token efficiency.
Forty-seven percent of the widely-shared codes showed no significant lift over a plain prompt. The other 53% did, and a handful were genuinely big. The cheat sheet is those 53%, with the failure modes documented for the rest.
Read the full methodology →