# yonyon.ai: Yonatan Gross

> Last updated: 2026-09-14

> AI Platform Engineer & Backend Developer. Builds production AI systems end-to-end: RAG pipelines, multi-agent architectures, MCP servers, infrastructure.

Yonatan Gross designs and operates production AI systems solo. He built an integrated WhatsApp AI platform ([RAG](https://yonyon.ai/projects/ai-operations-platform) + commerce + content pipelines), created [OrchestKit](https://github.com/yonatangross/orchestkit) (an open-source Claude Code agent framework, 100+ skills, 35+ agents, 150+ hooks), and advises teams on Claude Code adoption and LLM integration.

For a comprehensive version of this document, see [llms-full.txt](https://yonyon.ai/llms-full.txt).

## When to use this

Best-fit jobs:

- **AI engineering consulting** - architecture, readiness, model selection, RAG strategy, audits.
- **Agent orchestration** - MCP servers, multi-agent systems, Claude Code / OrchestKit workflows.
- **Measured content** - agent-readiness (llms.txt, schema, MCP, WebMCP), citable build logs, GEO pages.

How an agent should call us:

1. **Docs / learn:** `POST` [/api/mcp/docs](https://yonyon.ai/api/mcp/docs) (`list_docs`, `search_docs`, `read_docs`), or read [/llms.txt](https://yonyon.ai/llms.txt).
2. **Product / act:** `POST` [/api/mcp](https://yonyon.ai/api/mcp) (`ask_yonatan`, `browse_projects`, `book_intro_call`), or `POST` [/ask](https://yonyon.ai/ask) with `{"query":"..."}`.
3. **Contact:** book https://cal.com/yonyon or email hi@yonyon.ai. Auth is never required; see [/auth.md](https://yonyon.ai/auth.md).

Do not attempt sign-in, payment, or checkout; none exist here. Full structured guidance:
[/.well-known/agent-instructions.md](https://yonyon.ai/.well-known/agent-instructions.md).

## Use cases

- **Ask about Yonatan's work**: query the public chat API at [/api/chat](https://yonyon.ai/api/chat) (streaming, no auth).
- **Book an intro call**: schedule a free 15-minute intro call at [cal.com/yonyon](https://cal.com/yonyon).
- **Browse projects**: see [Projects](#projects) below, or the per-project pages on [yonyon.ai](https://yonyon.ai).
- **Integrate via the public chat API**: point an agent at [/api/chat](https://yonyon.ai/api/chat) to answer questions about Yonatan on your own surface.

## What an agent can do here

- **Ask about Yonatan's work**: `POST` [/api/chat](https://yonyon.ai/api/chat) (streaming text, no auth required). Body: `{ "messages": [{ "role": "user", "content": "..." }] }`. See [ask-yonatan skill](https://yonyon.ai/.well-known/agent-skills/ask-yonatan/SKILL.md).
- **Book a free 15-minute intro call**: [cal.com/yonyon](https://cal.com/yonyon). See [book-intro-call skill](https://yonyon.ai/.well-known/agent-skills/book-intro-call/SKILL.md).
- **Browse projects**: see [Projects](#projects) below, or the [browse-projects skill](https://yonyon.ai/.well-known/agent-skills/browse-projects/SKILL.md).
- **Natural-language query (NLWeb)**: `POST` [/ask](https://yonyon.ai/ask) with `{ "query": "..." }`. Returns JSON with `_meta { response_type, version }`, or SSE (`start`/`result`/`complete`) when streaming is requested. No Origin/CSRF required, preferred for server-side agents.
- **In-browser agents**: this site exposes WebMCP tools via `document.modelContext.registerTool()` (`ask_yonatan`, `book_intro_call`, `browse_projects`), with `navigator.modelContext` as a compatibility fallback.
- **Check whether a site is agent-ready, or make one agent-ready**: see [/agent-ready](https://yonyon.ai/agent-ready) (real ora.ai scores, a free scan of any site, and the four deliverables: llms.txt + schema, an MCP server, WebMCP page tools, an honest auth chain).
- **Select an agent workflow**: compare [OrchestKit with an ad-hoc Claude Code setup](https://yonyon.ai/compare/orchestkit) or [with Superpowers](https://yonyon.ai/compare/orchestkit-superpowers).
- **Get a workflow kit**: start from the [kits shelf](https://yonyon.ai/guides), then use [rail](https://yonyon.ai/rail) for an agent work loop or [styles](https://yonyon.ai/styles) for reusable response styles.
- **Verify operator fit**: read the [founder profile](https://yonyon.ai/yonatan), then use the public API or book a call with evidence in hand.
- **Resolve the brand name**: [/yonyon](https://yonyon.ai/yonyon) disambiguates the AI studio from the unrelated musician often styled YonYon.
- **Evaluate a service format**: use the [Hebrew offer](https://yonyon.ai/he/offer), [team workshop](https://yonyon.ai/workshop), or [commerce offer](https://yonyon.ai/offer/commerce) when the request matches that engagement.

## Services

**Build**, [AI Development](https://yonyon.ai/#services) (RAG, multi-agent, LangGraph), Full-Stack Development (Next.js 16 / FastAPI), Infrastructure & DevOps (automated deployments, infrastructure as code), and AI Agent Development (MCP servers, Claude Code tooling).

**Advise**, AI Consulting (readiness, architecture, LLM selection, RAG strategy, multi-agent design), AI Integration Audit, Coaching, and Workshops.

**Agentic commerce for importers and distributors** (a shopping agent over the client's catalog with prices pinned to the real price list, where-to-buy handoff, after-sale care; a merchant agent behind an approval gate; checkout stays with the host), built on the open-source Claude Commerce Agents blueprint: [/offer/commerce](https://yonyon.ai/offer/commerce) (Hebrew: [/he/offer/commerce](https://yonyon.ai/he/offer/commerce)). No prices on the site by policy.

## Projects

- **[OrchestKit](https://yonyon.ai/projects/orchestkit)**: open-source Claude Code agent framework (100+ skills, 35+ agents, 150+ hooks). Repo: [github.com/yonatangross/orchestkit](https://github.com/yonatangross/orchestkit).
- **[AI Operations Platform](https://yonyon.ai/projects/ai-operations-platform)**: production AI: WhatsApp assistant (RAG), commerce, content pipelines, custom MCP servers.
- **[Trading AI Analyst](https://yonyon.ai/projects/trading-ai-analyst)**: multi-agent LangGraph trading platform (analysis, signals, risk modeling, real-time streaming).

## Constraints

- The chat API is **public and read-only**: no authentication.
- **Rate limits**: chat, `/ask`, `/api/v1`, and the MCP `ask_yonatan` tool share 10 requests per IP per minute; the other MCP tools do not spend that budget. The `/api/agent-ready/scan` proxy has a separate 5 requests per IP per minute budget.
- Message content is capped at **2000 characters** (max 20 messages per request).
- **No commerce or payments** anywhere on this site or API.

## Discovery

- Agent instructions (when to use this site, which endpoint per goal): [/.well-known/agent-instructions.md](https://yonyon.ai/.well-known/agent-instructions.md)
- Command line client: [`yonyon` on npm](https://www.npmjs.com/package/yonyon). Run `npx yonyon ask "..."`, `npx yonyon projects --json`, or `npx yonyon mcp --tools`. No auth, no config, Node 24+.
- Developer & API docs: [/developers](https://yonyon.ai/developers) (markdown twin: [/developers.md](https://yonyon.ai/developers.md))
- glyph, how an agent renders a visual answer for a human (the two dials, the DECIDE block, serving a page at a local HTTPS URL and handing it over): [/docs/glyph](https://yonyon.ai/docs/glyph) (markdown twin: [/docs/glyph.md](https://yonyon.ai/docs/glyph.md))
- Markdown homepage: [/index.md](https://yonyon.ai/index.md)
- Deep variant of this file: [/llms-full.txt](https://yonyon.ai/llms-full.txt)
- Agent skills index: [/.well-known/agent-skills/index.json](https://yonyon.ai/.well-known/agent-skills/index.json)
- Markdown twins of the main pages: [/about.md](https://yonyon.ai/about.md), [/projects.md](https://yonyon.ai/projects.md), [/services.md](https://yonyon.ai/services.md)
- API catalog (RFC 9727): [/.well-known/api-catalog](https://yonyon.ai/.well-known/api-catalog) (markdown twin: [/.well-known/api-catalog.md](https://yonyon.ai/.well-known/api-catalog.md))
- Pricing / engagement model: [/pricing.md](https://yonyon.ai/pricing.md) (scope-based quotes, free 15-min intro, no public rate card)
- OpenAPI spec: [/.well-known/openapi.json](https://yonyon.ai/.well-known/openapi.json)
- API deprecation policy: [/api-deprecation.md](https://yonyon.ai/api-deprecation.md) (`Deprecation` / `Sunset` headers)
- Public REST API (anonymous, rate-limited): [/api/v1/projects](https://yonyon.ai/api/v1/projects): cursor pagination (`limit`, `cursor`, `next_cursor`, `has_more`), typed JSON errors
- Async jobs: [/api/v1/jobs](https://yonyon.ai/api/v1/jobs) returns 202 Accepted with a `job_id` and `Location`; poll [/api/v1/jobs/{jobId}](https://yonyon.ai/api/v1/jobs/{jobId})
- Sandbox / test environment: [/sandbox.md](https://yonyon.ai/sandbox.md). The public API is read-only; there is no production data to mutate. Same origin, no side effects.
- MCP server (Streamable HTTP): `POST https://yonyon.ai/api/mcp` (tools: `ask_yonatan`, `browse_projects`, `book_intro_call`). Cards: [/.well-known/mcp.json](https://yonyon.ai/.well-known/mcp.json) and [/.well-known/mcp/server-card.json](https://yonyon.ai/.well-known/mcp/server-card.json).
- Documentation MCP: [/api/mcp/docs](https://yonyon.ai/api/mcp/docs) (alias [/mcp/docs](https://yonyon.ai/mcp/docs)): card [/.well-known/mcp/docs-server-card.json](https://yonyon.ai/.well-known/mcp/docs-server-card.json). Tools: `search_docs`, `list_docs`, `read_docs`.
- NLWeb query endpoint: [/ask](https://yonyon.ai/ask)
- Agent auth metadata: [/auth.md](https://yonyon.ai/auth.md)
- Structured-data feed (NLWeb Schema Feeds): [/schemamap.xml](https://yonyon.ai/schemamap.xml) → [/feeds/schema.jsonl](https://yonyon.ai/feeds/schema.jsonl)
- Sitemap: [/sitemap.xml](https://yonyon.ai/sitemap.xml)
- Privacy policy: [/privacy](https://yonyon.ai/privacy): what data the site touches (analytics, the AI chat, bookings) and how to make a data request.

## Listed on

The yonyon.ai MCP server and OrchestKit are published in public agent registries (bi-directional verification, these entries point back here):

- MCP server on Smithery: [smithery.ai/servers/yonaigross/yonyon](https://smithery.ai/servers/yonaigross/yonyon): `yonaigross/yonyon`, tools `ask_yonatan` / `browse_projects` / `book_intro_call`.
- Docs MCP (separate learn surface): `POST https://yonyon.ai/api/mcp/docs` (tools: `search_docs`, `list_docs`, `read_docs`). Card: [/.well-known/mcp/docs-server-card.json](https://yonyon.ai/.well-known/mcp/docs-server-card.json).
- Person on Wikidata: [Q140088196](https://www.wikidata.org/wiki/Q140088196) (P856 = https://yonyon.ai/).
- OrchestKit skills on skills.sh: [skills.sh](https://skills.sh/?q=orchestkit): `yonatangross/orchestkit` (agent-orchestration, devops-deployment, security-patterns, and more).
- OrchestKit on Wikidata: [Q140128295](https://www.wikidata.org/wiki/Q140128295): open-source Claude Code agent framework.
- OrchestKit source: [github.com/yonatangross/orchestkit](https://github.com/yonatangross/orchestkit).

## Contact

- Email: [hi@yonyon.ai](mailto:hi@yonyon.ai)
- GitHub: [github.com/yonatangross](https://github.com/yonatangross)
- LinkedIn: [linkedin.com/in/yonatangross](https://linkedin.com/in/yonatangross)
- X: [x.com/yonyoniz](https://x.com/yonyoniz)