Documentation

Nexivo AI is a spec-driven software development platform. You describe what you want to build in plain English; the platform produces four engineering artifacts — Constitution, Specification, Architecture and Tasks — versioned and ready to ship to a delivery team.

This documentation walks through every surface of the product: the dashboard, the REST API, exports, webhooks and the credit model.

1. Receive a passcode from your administrator and sign in at /login.
2. Open the dashboard, paste an idea (one paragraph is enough), pick a stack hint and run the pipeline.
3. Review the four artifacts, regenerate any stage in place, and export the run to PDF, Markdown or JSON.

The dashboard uses passcode-based session authentication scoped to a workspace. API access uses bearer tokens issued from the Agency dashboard. Tokens inherit the credits and plan of the issuing workspace and can be scoped and revoked at any time.

Authorization: Bearer nxv_live_••••

• Constitution — non-negotiables, principles, success criteria.
• Specification — product scope, user stories, edge cases.
• Architecture — system diagram, data model, integration plan.
• Tasks — developer-ready breakdown for Jira, Linear and ClickUp.

Full endpoint reference lives on the API page. The base URL is https://api.nexivollc.com/v1. Every response uses standard HTTP status codes and an idempotency key header is supported on POST endpoints.

Official TypeScript and Python clients wrap the REST API with typed responses and retry handling.

import { Nexivo } from "@nexivo/sdk";
const nx = new Nexivo({ apiKey: process.env.NEXIVO_API_KEY });
const run = await nx.pipelines.create({ idea: "A workout tracker for climbers" });

• Generate a PRD nightly for the top-voted feature request.
• Hook a Slack /spec command into the pipeline.
• Fan a single idea into per-team architecture variants.

One pipeline run consumes one credit. Canceled runs do not consume credits. Credits refill on each successful billing cycle and roll back to zero on renewal. Inspect remaining credits via the dashboard header or the GET /v1/credits endpoint.

Every run is persisted as a versioned project. The Version History view groups runs by prompt with v1, v2, v3 lineage so you can regenerate any stage without losing prior outputs.

Nexivo routes requests across multiple upstream model providers with a zero-retention contractual posture wherever supported. Model selection is opaque by default; Agency customers can pin a provider per workspace.

The pricing endpoint returns the current plan catalog with localized currencies and credit allowances. Useful when rendering a pricing table inside an embedded onboarding flow.

Webhook payloads are signed with HMAC-SHA256. Verify the X-Nexivo-Signature header against the raw request body using your endpoint secret. Events: pipeline.completed, pipeline.failed, credit.exhausted, subscription.updated.

Default limits: 60 requests/min on Pro, 600 requests/min on Agency. Bursts up to 2x the limit are accepted for 10 seconds. Limits are returned in X-RateLimit-* response headers. Exceeded limits return 429 Too Many Requests with a Retry-After header.

• 400 — validation failed; the response includes a field-level errors object.
• 401 — missing or invalid bearer token.
• 402 — workspace is out of credits.
• 409 — conflicting state, often an in-flight run.
• 5xx — upstream failure; safe to retry with the same idempotency key.

Common questions are answered on the home page FAQ and in /faq.