Onairos - Unified User Data API for AI Agents
A user-owned data wallet giving AI agents rich persona, memory, and preferences across 1000+ OAuth sources — privacy-first, SDK-ready.
TL;DR
TL;DR: Onairos gives AI agents a privacy-first, user-owned data layer — persona, memory, and preferences pulled from 1000+ OAuth-connected sources via a single SDK.
Source and Accuracy Notes
This section is MANDATORY. All links must be verified from actual source, not guessed.
- Project page: onairos.io — verified product description
- Docs: docs.onairos.io/quick-setup — verified SDK install and setup steps
- SDK: npmjs.com/package/onairos — verified package
- HN launch thread: news.ycombinator.com/item?id=47070890
- Source last checked: 2026-07-29
What Is Onairos?
Onairos is a unified user data API built for AI agents. Instead of an AI agent starting cold with no context about a user, Onairos surfaces a rich persona — interests, memory, preferences — pulled from sources the user has already connected via OAuth. Think of it as a user-owned data wallet that apps can query with explicit consent.
The core promise, quoted directly from the site:
“Onairos turns your data into a personal model you own, control, and carry into your favorite apps.”
The product targets two audiences: AI app builders who want personalization without building their own data infrastructure, and end users who want control over what context they share with each app.
Core Capabilities
Onairos organizes around three pillars:
- Persona SDK — delivers a structured user profile (interests, sentiment, memory) to any LLM call in the first session, solving the cold-start problem
- Human API — a REST endpoint that enriches LLM calls with user context on demand
- Data Connectors — OAuth-based integrations with 1000+ sources: YouTube, Spotify, Reddit, LinkedIn, Netflix, ChatGPT, Claude, Gemini, Grok, Pinterest, Instagram, Sephora, and more
The SDK covers Web, React Native, Flutter, and SwiftUI with the same context payload across all platforms.
Setup Workflow
Step 1: Install the SDK
Onairos is published as an npm package:
npm install onairos
Platform-specific guides are available for React Native, Flutter, and Swift in the docs.
Step 2: Get an API Key
Sign up at developer.onairos.io to get a free API key. The free tier includes access to the Persona SDK and a set of core connectors.
Step 3: Configure with an AI Agent Prompt
Onairos ships with an “AI-assisted setup” flow. The setup prompt, as documented in the quick-start guide:
Run npm install onairos, then read node_modules/onairos/llm.txt and set up the Onairos web SDK with API key {Your_API_Key}.
After installing, the SDK reads its own llm.txt for platform-specific wiring instructions, then the AI agent handles the rest of the integration automatically.
Step 4: Connect Data Sources
Users connect their accounts through an OAuth consent flow embedded in the SDK. Each connector requires explicit user approval, and access can be revoked at any time. The connected sources populate the persona object returned on every SDK call.
Deeper Analysis
Privacy-First Architecture
Onairos stores data in a “user-owned wallet” — the user’s account, not Onairos’s servers. Data is only shared with apps when the user explicitly consents per-connection. The privacy policy states data is never resold, and GDPR/CCPA compliance is built into the consent model.
The architecture note from their docs: “User data ownership — data stays in the user’s wallet, never resold.”
Cold-Start Elimination
A persistent pain point for AI agents is first-session personalization. Without prior interaction history, agents default to generic responses. Onairos solves this by revealing a user’s already-expressed interests (from Reddit saves, YouTube watch history, Spotify likes, etc.) in the first SDK call — skipping the usual weeks of data collection.
Composio Complementary Positioning
Onairos explicitly positions alongside Composio: where Composio provides an actions layer (tool use, API calls), Onairos provides the context layer (who is the user, what do they care about). Apps can use both in combination.
Supported Integrations
Verified OAuth connectors (from the product schema): YouTube, Spotify, Reddit, LinkedIn, Netflix, ChatGPT, Claude, Gemini, Grok, Pinterest, Instagram, Sephora.
Practical Evaluation Checklist
- Free tier available at developer.onairos.io
- Web SDK:
npm install onairos— verified on npm - React Native, Flutter, SwiftUI SDKs available
- Data connectors: 1000+ sources via OAuth
- Privacy: GDPR/CCPA compliant, user-owned wallet
- Cold-start: First-session persona delivery confirmed in docs
- SDK version 8.3.3 (verified from product page schema)
Security Notes
- OAuth flows are user-initiated with per-app granular scopes
- Access can be revoked at any time by the user
- No passwords or private messages are collected
- On-device privacy-preserving inference is listed as a roadmap item (MIND 1)
FAQ
Q: Does Onairos work with self-hosted LLMs? A: Yes. The Human API and Persona SDK are HTTP-based and work with any LLM provider (OpenAI, Anthropic, local models) as long as the SDK call is made server-side or client-side with the API key.
Q: Is there a GitHub repository? A: Onairos does not appear to have a public source repository. The SDK is distributed via npm and the docs are at docs.onairos.io. This is a hosted/API product rather than open-source software.
Q: How is this different from Google Login or Apple Sign-In? A: Those services share only authentication (who you are). Onairos shares rich context about the user — interests, preferences, memory — with the user’s explicit permission per app.
Q: What happens if a user revokes access? A: The connected source immediately stops returning data to the app. The persona object updates to reflect the removed connector.
Conclusion
Onairos solves a real gap in the AI agent stack: context. Instead of every AI app rebuilding user profile infrastructure, Onairos provides a shared, consent-based data layer that agents can query on day one. The OAuth connector model means users bring their own data, and the privacy-first wallet means they control what flows where.
If you are building an AI app that would benefit from knowing your users faster — not just authenticating them — Onairos is worth a look. The free tier is sufficient for development and small production workloads.
Get started at developer.onairos.io or read the quick-start at docs.onairos.io/quick-setup.
Related Posts
ai-setup
Recall – Persistent Memory for Claude Code via MCP Hooks
Recall gives Claude Code a permanent memory store that survives session restarts and context compaction. Four hooks capture and restore context automatically — with cloud SaaS or self-hosted options.
2/28/2026
dev-tools
Automotive Skills Suite for AI Engineering
Evaluate Automotive Skills Suite for APQP, ASPICE, HARA, safety-plan, and DIA workflows with setup notes, governance risks, and SME review guidance.
5/28/2026
dev-tools
awesome-agentic-ai-zh Roadmap Guide
Explore awesome-agentic-ai-zh as a Chinese agentic AI learning roadmap, with setup notes, track selection, study workflow, and evaluation guidance.
5/28/2026