Skip to content
View ferhatatagun's full-sized avatar
14 followers · 23 following

Achievements

Achievement: Pair Extraordinairex4Achievement: Pull Sharkx3Achievement: YOLOAchievement: Quickdraw

Achievements

Achievement: Pair Extraordinairex4Achievement: Pull Sharkx3Achievement: YOLOAchievement: Quickdraw

Block or report ferhatatagun

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ferhatatagun/README.md

╭─ ferhat@atagun ───────────────────────────────────────────╮
│ Frontend Team Lead · AI-Native Frontend Engineer │
│ Istanbul · HangiKredi · open to AI-native frontend work │
╰───────────────────────────────────────────────────────────╯

site blog dev.to linkedin x

$ whoami

Frontend team lead in Istanbul, at HangiKredi. I build interfaces for AI — turning model behaviour into something you can see and reason about.

Lately I've been shipping a small suite of browser-only developer tools that make the Anthropic Claude API legible: prompt caching, token economics, agent tool-loops, side-by-side prompt A/B, and pre-flight token counting. All BYOK, no backend, source open.

$ ls -la /tools

claudoscope

claudoscope  · live ↗

X-ray your Claude API calls — prompt caching, token composition and cost, visualized live as the response streams.

 agent-replay

agent-replay  · live ↗

Replay a Claude agent's tool-calling loop step by step on a cinematic timeline.
prompt-lab

prompt-lab  · live ↗

A/B test prompts side by side — one input, two variants, compared on cost and latency.

 tool-lab

tool-lab  · live ↗

Interactive Claude tool-use sandbox — define tools, mock the responses, watch the agent loop live.
context-lens

context-lens  · live ↗

See a Claude prompt before you ship it — token breakdown, context-window position, cost estimate, prompt-caching boundaries. Pre-flight to claudoscope's post-hoc x-ray.

Long-form posts on each tool: ferhatatagun.com/blog · Suite overview: ferhatatagun.com/tools

$ cat .stack

languages:   TypeScript, JavaScript
frameworks:  React, Next.js, SvelteKit
styling:     Tailwind, Framer Motion
runtime:     Node.js, edge functions
api:         Anthropic Claude · Server-Sent Events · tool-use
deploy:      Vercel, GitHub Pages

$ cat now.txt

Currently shipping: context-lens — pre-flight token counter for Claude prompts. Writing about: dev tools, LLM observability, BYOK patterns. Reading on: ferhatatagun.com/blog · dev.to/ferhatatagun

Istanbul, Türkiye · open to interesting AI-native frontend work.

Pinned Loading

  1. prompt-lab prompt-lab Public

    A/B test your Claude prompts side by side — run one input through two variants at once, compare output, cost and latency.

    TypeScript 2

  2. agent-replay agent-replay Public

    Watch a Claude agent's tool-calling loop — paste a trace, replay it step by step on a cinematic timeline. No key, no backend.

    TypeScript 3

  3. expressjs-omdbapi expressjs-omdbapi Public

    A sample project created using the data that omdbapi provides

    SCSS 2

  4. claudoscope claudoscope Public

    See through to what Claude is doing — a BYOK Anthropic API playground that visualizes prompt caching, tokens, and cost in real time.

    TypeScript 2

  5. nextjs-pollsapi-apiary-app nextjs-pollsapi-apiary-app Public

    Crud operations with poll-api apiary in nextjs.

    JavaScript 2

  6. begin-react-native-github-api begin-react-native-github-api Public

    React Native Github Activity Feed example application

    Makefile 2