Stream a reply from the company assistant (xAI Grok). Server-Sent Events using the AG-UI event format.
10 / hour / IP · sliding windowStreaming
Developer & Agent Resources
We don't pitch what we haven't already built.
Select Interactive builds AI-ready, agent-friendly web platforms for our clients. This page is the public surface of our own site, exposed the same way we would expose yours: an OpenAPI specification, an MCP server, a published llms.txt, and content endpoints documented in OpenAPI 3.1 and exposed over MCP.
Public APIs
Six endpoints. Full OpenAPI spec. Documented for agents.
Every endpoint below is documented in our OpenAPI 3.1 specification. All endpoints are JSON over HTTPS and rate-limited per IP.
Submit a project inquiry. Persists to Firestore, notifies Slack, emails the team via SendGrid.
Per-submission abuse checks
Submit a project blueprint intake. Streams a templated, AI-generated overview back to the caller.
Per-IP and per-email limitStreaming
Counts of published case studies, news articles, and services on the public site.
Cached JSON snapshot
Liveness check. Add ?deep=1 with a bearer token for a Firestore-aware deep probe.
Liveness probe
Build identifiers (semver, git commit, deploy title, deploy time).
Build & deploy identifiers
JSON over HTTPSRate-limited per IP
Agent Surfaces
Discovery endpoints for AI crawlers and agents.
Six well-known surfaces that let machines find, understand, and interact with the public site, without scraping HTML.
01·Specifications
Machine-readable manifests.
Authoritative schemas for the API surface and the public content tree.
2 surfaces
OpenAPI 3.1 specification for every endpoint listed below. Suitable for code generation and SDK tooling.
Consumed by
- openapi-generator
- Stainless
- Swagger UI
Standard XML sitemap with every public page and content item.
Consumed by
- Googlebot
- Bingbot
- DuckDuckBot
02·Crawl Directives
Rules and a reading list for bots.
Plain-text instructions and a curated map for the next generation of AI crawlers.
2 surfaces
Plain-text site map authored for LLMs and AI search crawlers. Lists services, news, and key pages.
Consumed by
- ChatGPT Search
- Claude
- Perplexity
Crawl directives. Major AI crawlers are explicitly allowed.
Allow-listed crawlers
- GPTBot
- ClaudeBot
- PerplexityBot
- OAI-SearchBot
03·Agent Protocols
Interactive surface for autonomous agents.
Discoverable Model Context Protocol endpoint, agents can call tools, not just read pages.
2 surfaces
Model Context Protocol server (Streamable HTTP transport) exposing read-only tools for case studies, services, news, and the company assistant.
Consumed by
- Claude Desktop
- Cursor
- MCP Inspector
MCP discovery manifest. Points agents to the live MCP endpoint and lists its capabilities.
Consumed by
- MCP clients
- Agent runtimes
Usage Examples
Get started in under a minute.
Ask the assistant
curl -N -X POST https://www.select-interactive.com/api/ask \
-H 'Content-Type: application/json' \
-d '{
"messages": [{
"id": "1",
"role": "user",
"parts": [{ "type": "text", "content": "What services does Select Interactive offer?" }]
}]
}'Read content statistics
curl https://www.select-interactive.com/api/content-statsStatus & Versioning
Always-on health and build metadata.
Two unauthenticated endpoints for monitoring and release tracking. Backwards-incompatible response changes are flagged via the version field, which follows semver across major releases.
Liveness check. Add ?deep=1 with a bearer token for a Firestore-aware deep probe.
{
"status": "ok",
"uptime": "27d 14h"
}Liveness probe
Current build identifiers: semver, git commit, deploy title, and deploy time.
{
"version": "1.1.9",
"commit": "9667f64",
"deployedAt": "2026-05-10T18:24:00Z"
}Semver · follows major releases
Build With Us
Want this for your product?
We design and build agent-ready web platforms, OpenAPI surfaces, and Model Context Protocol servers for clients. If your team is making your product accessible to AI agents and wants a partner who has done it on their own site first, get in touch.