VIKKA / DEVELOPER RESOURCES
Vikka Studio API and developer resources.
Vikka is a project-based web design and development studio, not a self-service software platform. These resources help developers and AI agents verify what Vikka offers, read the site without rendering its interface, inspect public work, and prepare an accurate project enquiry.
01
Hosting and implementation
The Vikka website is built with Next.js and deployed on Vercel. Canonical content URLs support HTTP content negotiation: clients that prefer text/markdown can request the same public URL with an Accept header and receive clean Markdown with a matching Vary header.
The sitemap lists indexable pages, while llms.txt provides a curated entry point and specific when-to-use guidance. Agent instructions describe suitable projects, unsupported claims, and the safest way to prepare a contact request.
02
Vikka Agent Resources API
Vikka exposes a small public JSON API for accurate discovery. GET /api/v1/services lists the current project types, GET /api/v1/projects returns the selected work Vikka publicly presents, and POST /api/v1/project-brief validates business context and prepares reviewable contact links.
These endpoints require no authentication because they expose public information and have no sending, purchasing, booking, or publishing side effects. Vikka does not currently offer webhooks or a client-management API and does not currently publish an official CLI. Scripts should use the documented HTTP API, OpenAPI contract, or MCP endpoint.
03
OpenAPI 3.1 and JSON errors
The contract is published at /openapi.json using OpenAPI 3.1. Every operation has a unique operationId, a plain-language description, typed input and output schemas, and documented error responses so an agent can translate it into a function call without guessing.
Every /api path returns JSON, including missing routes, unsupported methods, malformed request bodies, and validation failures. Error responses include a stable code, a readable message, a recovery hint, and a link back to the OpenAPI specification.
04
Rate limits and retries
The public API allows 60 requests per client in a 60-second window. Every response publishes RateLimit-Policy and RateLimit, together with the widely supported RateLimit-Limit, RateLimit-Remaining, and RateLimit-Reset compatibility fields.
When the quota is exceeded, Vikka returns HTTP 429 with the same structured JSON error shape and a Retry-After header. Clients should wait for Retry-After before retrying instead of guessing or creating rapid retry loops.
05
API versioning and deprecation
Stable resources use a major version in the URL, such as /api/v1. The policy at /api-versioning.md explains compatible additions, breaking changes, and a target notice period of at least 180 days.
API v1 is active and is not deprecated. If that changes, Vikka will use the standards-based Deprecation header and migration link; a Sunset header will only appear when an actual shutdown date has been scheduled.
06
Model Context Protocol
The first-party Model Context Protocol endpoint lives at /.well-known/mcp and uses Streamable HTTP. Its tools retrieve Vikka services and selected work or prepare a project brief and contact links. Preparing a brief does not send a message, create a contract, reserve availability, or produce a price.