Console SDK

Cognipeer Console SDK

TypeScript for the Cognipeer Console. One client, every Console resource — chat, agents, browsers, vectors, files, tracing. Calm types, streaming-first, OpenAI-compatible where it matters.

Package @cognipeer/console-sdk
Language TypeScript, no any escape hatches
API OpenAI-compatible where it matters

The SDK is the application-layer client; Console is the platform underneath it. Reach for the SDK when you're wiring Cognipeer into a Node, Next.js, or browser app and want resource methods instead of raw HTTP. Go to the Console docs when you're provisioning tenants, providers, or deployment targets instead.

typescript

import { ConsoleClient } from '@cognipeer/console-sdk';

const client = new ConsoleClient({ apiKey: process.env.COGNIPEER_API_KEY! });

const stream = await client.chat.completions.create({
  model: 'gpt-4o',
  stream: true,
  messages: [
    { role: 'system', content: 'You are a precise assistant.' },
    { role: 'user',   content: 'Summarize the Q3 release notes.' },
  ],
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content ?? '');
}

Guide

Getting started

Install, authenticate, and send your first request.

Guide

Working with Console

How platform ops and app integration divide responsibility.

Reference

Console API mapping

A one-to-one method-to-endpoint reference.

Reference

API reference

Every resource on the client, from chat to sandbox.

Examples

Chat, streaming, memory, tracing, and reranking — runnable.

Project

Release notes

What shipped, release by release.

Every Console resource, one client

Resource	What it gives you	Reference
Chat completions	OpenAI-compatible chat with first-class streaming — drop it in where you already call OpenAI.	Chat
Agents API	Multi-turn Responses with tool bindings, version pinning, and structured outputs.	Agents
Unified tools	List, execute, and convert OpenAPI and MCP tools into Agent SDK objects.	Tools
Browser automation	Create browser profiles, drive live sessions, expose each as an MCP endpoint.	Browser
Embeddings & vectors	One API across Chroma, Postgres, Elasticsearch, Azure AI Search, and S3 Vectors.	Vectors
Files & memory	Upload documents with markdown conversion, thread them into long-lived memory.	Files
Config & guardrails	Manage secrets and group resolution, apply guardrails per environment.	Guardrails
Tracing	Every agent run produces a structured trace — spans, tool calls, latencies.	Tracing
Realtime & voice	WebSocket sessions with streaming responses and an optional STT/TTS round-trip.	Realtime
Agent Sandbox	Remote runtime containers — exec, run code, manage files and git, snapshot or fork.	Sandbox
Batch, moderations & spend	Async batch inference, guardrail-backed moderation, spend reports with budget caps.	Batches
Strict types	Full TypeScript, no `any` escape hatches — generics carry through streams and tools.	Type safety

Where to go next

Getting started — install, auth, base URL, the first request.
Working with Console — the split between platform ops and app integration.
Console API mapping — every HTTP endpoint, mapped to its SDK method.
Examples — chat, streaming, RAG, memory, and tracing, end to end.

Stuck somewhere? Open a GitHub issue — the team triages from there. For platform questions (provider keys, tenants, billing) head to the Console docs instead.

Cognipeer Console SDK ​