wansify.venturesScreenshots
Product details
Shurikly is a free, open-source (MIT) desktop API client for macOS, Windows and Linux. It sends REST, GraphQL and WebSocket requests, turns a project folder into a collection with its Code Scanner, generates and runs tests with its AI Engineer, and runs local mock servers — with no account and no cloud.
Backend and frontend developers, API teams and QA engineers who want a private, local-first API client that starts from their code instead of a blank request.
Point Shurikly at a project folder and it detects backend routes and frontend calls, then builds a ready-to-send collection. Send and inspect requests, compare responses, run scripts and tests, spin up mock servers, and optionally let the AI Engineer write tests using OpenAI, Anthropic, Gemini or a local Ollama model. Data stays on your machine as JSON files.
Features
Shurikly is a free, open-source (MIT) desktop application for designing, sending, testing, mocking, and documenting HTTP APIs on macOS, Windows, and Linux. It is local-first: there is no account, no cloud sync, and no telemetry — collections are stored as JSON files on your machine and API keys are encrypted with the operating system's secure storage. Instead of starting from a blank request, Shurikly starts from your code. Point it at a project folder and the Code Scanner finds the routes in your backend and the calls in your frontend, then builds a ready-to-send collection. The AI Engineer generates and runs tests using OpenAI, Anthropic, Gemini, or a local Ollama model. Version 1.0.0 launches on 25 October 2026.
Params, headers, JSON, form, raw and binary bodies, with Bearer, JWT, Basic, Digest, API key, OAuth 1.0 & 2.0, AWS Signature and Hawk auth.
Scans Express, Fastify, NestJS, Next.js, FastAPI, Flask, Django and OpenAPI projects, plus frontend fetch and axios calls, into a ready-to-send collection.
Generates and runs API tests with OpenAI, Anthropic, Gemini or local Ollama, and explains failures with evidence.
Local HTTP mock servers with multiple responses per route — active, sequential or random.
Model Context Protocol support so Claude and other AI agents can work with your collections.
Pre-request scripts, assertions, and a collection runner that produces readable reports.
Diff status, headers, JSON paths, text, timing and size against the last run, a pinned baseline or another tab.
Import from Postman, OpenAPI and cURL; export to Postman, OpenAPI, cURL, Markdown and HTML; generate JavaScript, Node.js and Python code.
Benefits
Clear outcomes from day one — productivity, clarity, and growth without the clutter.
The open-source API client that reads your code
Suitable for
Backend and frontend developers, API teams, QA engineers, and privacy-conscious organisations that want a local-first alternative to cloud API clients.
FAQ
Shurikly is a free, open-source desktop API client for REST, GraphQL and WebSocket that scans your code into collections, writes tests with built-in AI and runs mock servers, all locally.
Yes. It is MIT-licensed open source with no paid tier, no seat limits and no account required.
Version 1.0.0 launches on 25 October 2026 for macOS, Windows and Linux. Installers and the public GitHub repository go live on launch day.
No. Collections are stored as local JSON files, API keys are encrypted with the operating system's secure storage, and sending data to AI providers is off by default with secret redaction.
Yes. Shurikly imports from Postman, OpenAPI and cURL, and exports to Postman, OpenAPI, cURL, Markdown and HTML.
AI add-on
Every Wansify product can be extended with custom AI assistants and chatbots — trained on your data, embedded in your workflow, and ready to support customers or staff 24/7.
Extend Shurikly with an AI assistant trained on your business data — for customers, staff, or both.
In-app AI assistant inside CRM, LMS, gym app, store, or support desk.
Responses grounded in your manuals, FAQs, product catalog, and policies.
Floating chat on your site that connects to your product or knowledge base.
OpenAI, Claude, Gemini, or local models with custom prompts and guardrails.
