Pressline Developer Documentation

This technical guide documents the programmatic interfaces, Model Context Protocol (MCP) servers, and orchestrator execution parameters of the Pressline Autonomous Publishing Engine.

1. Quickstart & Trigger Execution

A generation and deployment cycle can be triggered via Node.js or via GitHub Actions webhooks:

// Trigger standard programmatic execution
const { runPipeline } = require('./automation/runner');

async function main() {
  const result = await runPipeline({
    autoDeploy: true,
    forceCategory: 'Commercial Real Estate & ROI'
  });
  console.log(`Published post: ${result.slug}`);
}

main();

2. Model Context Protocol (MCP) Server

Pressline provides an official MCP endpoint allowing AI agents (Claude, ChatGPT Operator, Cursor) to interface with the knowledge base and trigger publishing tools directly:

• Manifest Discovery Endpoint: https://pressline.vercel.app/.well-known/mcp.json
• Protocol: Model Context Protocol 2024-11-05 (Streamable HTTP)
• Available Tools:
  • pressline_query_knowledge_base: Semantic and factual lookup across ground truth registries.
  • pressline_audit_article: Run the multi-pass Critic v5 quality gatekeeper on arbitrary draft HTML.
  • pressline_generate_cover: Compile custom 16:9 branded SVG artwork for a given title and category.
  • pressline_run_pipeline: Trigger an end-to-end publish cycle with auto-deployment.

3. OpenAPI 3.1.0 Specification

Pressline publishes an official self-describing OpenAPI 3.1.0 specification with unique operationId attributes and full JSON Schema definitions on every endpoint:

• JSON OpenAPI Spec: https://pressline.vercel.app/openapi.json
• YAML OpenAPI Spec: https://pressline.vercel.app/openapi.yaml
• Well-Known Spec: https://pressline.vercel.app/.well-known/openapi.json

4. Official CLI Tool (`pressline`)

Developers and autonomous agents can script interactions using the official CLI package:

# Run via npx directly
npx pressline run

# Audit an article draft against Critic v5
npx pressline audit article-draft.html

# Query ground truth knowledge base
npx pressline query "Etihad Town Phase 4"

# Start local Model Context Protocol JSON-RPC stream
npx pressline mcp

5. Machine-Readable Agent Resources

AI agents and crawlers can consume our standardized machine-readable context files:

• LLMs.txt Index: https://pressline.vercel.app/llms.txt
• LLMs-Full Comprehensive Dump: https://pressline.vercel.app/llms-full.txt
• OpenAPI 3.1.0: https://pressline.vercel.app/openapi.json
• Agent Instructions: https://pressline.vercel.app/.well-known/agent-instructions.json
• MCP Tool Manifest: https://pressline.vercel.app/.well-known/mcp.json
• Sitemap XML: https://pressline.vercel.app/sitemap.xml

6. Content Negotiation (`Accept: text/markdown`)

Pressline fully complies with the acceptmarkdown.com standard. When requesting any public URL with an Accept: text/markdown header, our edge router returns pure markdown with a valid Vary: Accept, Accept-Encoding header.

curl -H "Accept: text/markdown" https://pressline.vercel.app/