bb77eb5f35
Re-license Pry from full Proprietary to a dual-license model: - Core engine, extraction, templates (80+), MCP server, x402 payment rail, CLI, SDK, browser extension, WordPress plugin, Shopify app, and llm_providers: MIT (see LICENSE) - Anti-detection / stealth subset (15 files): BSL 1.1 with Change Date 2029-01-01 (see LICENSE-BSL-STEALTH) BSL files (anti-detection moat): ultimate_scraper.py, stealth_engine.py, stealth_scripts/*.js (6), camoufox_integration.py, tls_fingerprint.py, cookie_warmer.py, behavioral_biometrics.py, adaptive.py, browser_pool.py, network.py, captcha_solver.py, shadow_dom.py, lazy_load.py, signup_automator.py, auth_connector.py This enables community contributions to the core engine (templates, integrations, MCP tools) while protecting the anti-detection techniques that constitute the actual competitive moat. BSL Additional Use Grant permits free non-production use; production deployment requires a commercial license from enterprise@rugmunch.io. Changes: - Replace proprietary LICENSE with MIT LICENSE + new LICENSE-BSL-STEALTH - Add SPDX-License-Identifier headers to 300+ source files - Add docs/adr/0002-dual-licensing.md (ADR documenting the decision) - Update README.md: new License section with BSL Additional Use Grant - Update LICENSING_PRICING_STRATEGY.md: Section 3 (PryScraper) for dual license - Update AGENTS.md: license line in header + new rule 8 (PRs touching BSL rejected) - Update pyproject.toml: license = "MIT AND BSL-1.1" + classifiers + license-files - Update DECISIONS.md index with ADR-0002 - Update STATUS.md (2026-07-03) and PLAN.md sprint goals Refs: ADR-0002
133 lines
4.9 KiB
Python
133 lines
4.9 KiB
Python
"""Pry MCP Server — expose scrape, crawl, automate as MCP tools.
|
|
Enables any MCP-compatible AI agent (Claude, Hermes, Cursor) to use Pry."""
|
|
|
|
# SPDX-License-Identifier: MIT
|
|
# Copyright (c) 2026 Rug Munch Media LLC
|
|
#
|
|
# Part of Pry — https://git.rugmunch.io/RugMunchMedia/pryscraper
|
|
# Licensed under MIT. See LICENSE.
|
|
|
|
import httpx
|
|
|
|
|
|
class PryMCPServer:
|
|
"""MCP server exposing Pry as tools for AI agents.
|
|
|
|
Implements the Model Context Protocol (MCP) for tool discovery.
|
|
"""
|
|
|
|
def __init__(self, base_url: str = "http://localhost:8002"):
|
|
self.base_url = base_url
|
|
self._client = httpx.Client(timeout=60)
|
|
|
|
def list_tools(self) -> list[dict]:
|
|
"""MCP tool discovery — returns available tools."""
|
|
return [
|
|
{
|
|
"name": "pry_scrape",
|
|
"description": "Scrape a URL to clean markdown. Bypasses Cloudflare automatically.",
|
|
"inputSchema": {
|
|
"type": "object",
|
|
"properties": {
|
|
"url": {"type": "string", "description": "URL to scrape"},
|
|
"timeout": {
|
|
"type": "integer",
|
|
"description": "Timeout in seconds",
|
|
"default": 30,
|
|
},
|
|
},
|
|
"required": ["url"],
|
|
},
|
|
},
|
|
{
|
|
"name": "pry_crawl",
|
|
"description": "Crawl multiple pages from a starting URL.",
|
|
"inputSchema": {
|
|
"type": "object",
|
|
"properties": {
|
|
"url": {"type": "string", "description": "Starting URL"},
|
|
"maxPages": {"type": "integer", "description": "Max pages", "default": 10},
|
|
},
|
|
"required": ["url"],
|
|
},
|
|
},
|
|
{
|
|
"name": "pry_map",
|
|
"description": "Discover all URLs on a website.",
|
|
"inputSchema": {
|
|
"type": "object",
|
|
"properties": {
|
|
"url": {"type": "string", "description": "Site URL"},
|
|
"limit": {"type": "integer", "description": "Max links", "default": 50},
|
|
},
|
|
"required": ["url"],
|
|
},
|
|
},
|
|
{
|
|
"name": "pry_parse",
|
|
"description": "Parse a document (PDF, DOCX, image) to text.",
|
|
"inputSchema": {
|
|
"type": "object",
|
|
"properties": {
|
|
"url": {"type": "string", "description": "Document URL"},
|
|
},
|
|
"required": ["url"],
|
|
},
|
|
},
|
|
{
|
|
"name": "pry_screenshot",
|
|
"description": "Take a screenshot of a URL.",
|
|
"inputSchema": {
|
|
"type": "object",
|
|
"properties": {
|
|
"url": {"type": "string", "description": "URL to screenshot"},
|
|
},
|
|
"required": ["url"],
|
|
},
|
|
},
|
|
{
|
|
"name": "pry_automate",
|
|
"description": "Execute browser automation steps (navigate, click, type, extract). Login automation, form filling.",
|
|
"inputSchema": {
|
|
"type": "object",
|
|
"properties": {
|
|
"steps": {
|
|
"type": "array",
|
|
"items": {"type": "object"},
|
|
"description": "List of automation steps. Each step: {action, selector?, value?, url?}",
|
|
}
|
|
},
|
|
"required": ["steps"],
|
|
},
|
|
},
|
|
]
|
|
|
|
async def call_tool(self, name: str, arguments: dict) -> dict:
|
|
"""Execute an MCP tool call."""
|
|
tool_map = {
|
|
"pry_scrape": (
|
|
"/v1/scrape",
|
|
{"url": arguments["url"], "timeout": arguments.get("timeout", 30)},
|
|
),
|
|
"pry_crawl": (
|
|
"/v1/crawl",
|
|
{"url": arguments["url"], "maxPages": arguments.get("maxPages", 10)},
|
|
),
|
|
"pry_map": (
|
|
"/v1/map",
|
|
{"url": arguments["url"], "limit": arguments.get("limit", 50)},
|
|
),
|
|
"pry_parse": ("/v1/parse", {"url": arguments["url"]}),
|
|
"pry_screenshot": ("/v1/screenshot", {"url": arguments["url"]}),
|
|
"pry_automate": ("/v1/automate", {"steps": arguments["steps"]}),
|
|
}
|
|
|
|
if name not in tool_map:
|
|
return {"error": f"Unknown tool: {name}"}
|
|
|
|
path, payload = tool_map[name]
|
|
async with httpx.AsyncClient(timeout=120) as client:
|
|
resp = await client.post(f"{self.base_url}{path}", json=payload)
|
|
return resp.json()
|
|
|
|
|