@ai11y/core

Framework-agnostic core. Your app exposes UI via markers (elements with data-ai-id, data-ai-label, optional data-ai-intent). The API is describe → plan → act.

API

• createClient(options?) — Returns { describe, act, track, reportError }. Options: onNavigate?: (route: string) => void.
• describe() — Returns Ai11yContext: markers, route, state, optional error. Scans the DOM for marked elements and which are in view.
• act(instruction) — Executes one instruction: click, navigate, highlight, scroll, or fillInput.
• plan(ui, input, config?, messages?) — Runs the agent (LLM or rule-based) and returns Promise<{ reply, instructions }>.

JavaScript example:

import { createClient, plan } from "@ai11y/core";

const client = createClient({
  onNavigate: (route) => window.history.pushState({}, "", route),
});

const ui = client.describe();
const { reply, instructions } = await plan(ui, "click the save button");
for (const instruction of instructions ?? []) {
  client.act(instruction);
}

Types (Ai11yContext, Instruction, AgentResponse, etc.) and DOM helpers (clickMarker, highlightMarker, navigateToRoute, etc.) are exported from the main entry. See generated docs for full reference.