ScrapeGraphAI/scrapegraph-py

ScrapeGraphAI SDK is a thin client wrapper around a proprietary SaaS API. The defensibility is critically weak for three reasons: (1) This is not the core technology—the value sits entirely in the backend service, which the user cannot control or replicate. (2) At 69 stars with 0 velocity, adoption is minimal and the project shows no momentum. (3) The SDK itself is a standard REST/HTTP client library with pydantic models—trivially reproducible boilerplate. Any competitor offering web scraping + LLM-based extraction (Bright Data, Apify, Firecrawl, or built-in capabilities from OpenAI/Anthropic) can replicate this surface. Platform domination risk is HIGH because: OpenAI, Anthropic, Google, and AWS are all adding native web scraping and data extraction to their AI platforms. Anthropic's recent computer-use models and OpenAI's browsing capabilities directly compete. Market consolidation risk is HIGH because the web scraping/data extraction market is crowded with well-funded incumbents (Apify $80M+, Bright Data, Selenium community, etc.) and LLM labs are rapidly commoditizing agentic browser automation. Displacement horizon is 6 months because LLM providers are actively shipping competitive features, and users will consolidate on platforms that offer scraping as a native capability rather than maintaining a separate SDK. The project has no community lock-in, no irreplaceable data, no switching costs, and no technical moat. It is entirely dependent on the backend service remaining differentiated and available.