# stacktxt — Full Technical Specification & Machine Knowledge Index

> The standard in agentic infrastructure, crawler directives, structured schema, and markdown documentation layers for modern commerce.

## System Architecture & Mission
stacktxt provides turn-key engineering infrastructure that bridges traditional human-optimized web applications with autonomous AI agents. Modern AI models (ChatGPT, Claude, Perplexity, Gemini, Apple Intelligence) do not navigate via heavy client-side JavaScript or bloated visual DOM trees. They consume semantic, high-density, low-token text representations.

stacktxt optimizes this interface across three foundational pillars:
1. Crawlability & Permission Controls (`robots.txt`)
2. Machine Knowledge & Navigation Layers (`/llms.txt`, `/llms-full.txt`)
3. Structured Entity Graphs (Schema.org JSON-LD & Protocol Manifests)

---

## Technical Capabilities & Core Deliverables

### 1. Crawlability & robots.txt Directive Engineering
- Full crawlability analysis of robots.txt across all major AI agent user-agents.
- Strict isolation of search/citation bots (`OAI-SearchBot`, `ChatGPT-User`, `PerplexityBot`, `Claude-SearchBot`, `Applebot-Extended`) versus scrapers and bulk training bots.
- Disallow protection on private administrative, checkout, and search-filter URL loops to eliminate crawl budget exhaustion.

### 2. Machine Reading Layer (/llms.txt & /llms-full.txt)
- High-density markdown documentation deployed at the domain root.
- Significant reduction in model context window token consumption compared to raw HTML scraping.
- Automated catalog and sitemap synchronization ensuring models reference live pricing, availability, and specs without hallucination.

### 3. Structured Entity Graphs (Schema.org JSON-LD)
- Canonical Organization, WebSite, and ProfessionalService entity graph disambiguation.
- Verification of sameAs social, business profile, and code repository links.
- Implementation of rich product, offer, return policy, and merchant metadata.

---

## Technical Checks & Evaluation Standards

### Crawlability Checks
- OAI-SearchBot: Must be explicitly permitted for real-time search discovery.
- ChatGPT-User: Must be permitted for user-prompted web browsing.
- PerplexityBot: Must be permitted with clean access to product documentation.
- Claude-SearchBot / ClaudeBot: Must be permitted for Anthropic answer generation.
- Applebot-Extended: Must be permitted for Apple Intelligence spotlight and Siri suggestions.

### Machine Reading Layer Checks
- `/llms.txt`: Must exist at domain root, serve HTTP 200 with text/markdown or text/plain, and remain concise.
- `/llms-full.txt`: Must provide an exhaustive, authoritative machine-readable technical index.
- Token Efficiency: Clean markdown layers reduce LLM context token consumption by 90-95% compared to raw HTML DOM trees.

### Entity Graph Checks
- Organization entity with verified sameAs social and institutional identifiers.
- Unambiguous canonical URLs on all schema nodes.
- Valid Schema.org JSON-LD without syntax errors or broken references.

---

## Machine Protocols & Agent Discovery
- WebMCP Manifest: https://stacktxt.com/.well-known/webmcp.json
- OpenAI Plugin Manifest: https://stacktxt.com/.well-known/ai-plugin.json
- Robots Directives: https://stacktxt.com/robots.txt
- Sitemaps: https://stacktxt.com/sitemap.xml

---

## Contact & Institutional Channels
- Maxwell Thomason, Co-Founder: max@stacktxt.com
- Kellen, Co-Founder: kellen@stacktxt.com
- Technical Inquiries: team@stacktxt.com
- Official Domain: https://stacktxt.com
- Booking Portal: https://stacktxt.com/book