nostrband/ServiceGraph: AI Agent Skills for Structured Business Data

What changed

nostrband has introduced ServiceGraph, a set of AI agent skills that allow artificial intelligence agents to interact with and extract information from structured business datasets. The project aims to provide startup founders and developers with actionable data for tasks such as identifying potential markets, finding contact information, and assessing company sizes. The ServiceGraph skill acts as a universal entry point, enabling agents to discover available datasets, understand their schemas, and perform searches. Specific skills are also available for more targeted queries. Currently, the available datasets include over 110,000 US professional-services firms (agencies) enriched with details on services, size, and location, and over 3,500 business directories containing industry, domain rating, and traffic information. Additionally, there are over 850 product directories for software launches and backlink opportunities, and over 50,000 newsletters with subscriber counts, topics, and posting cadences. Future datasets are planned for subreddits and influencers.

The find-service-providers skill, for instance, allows agents to filter US professional-services firms across 22 industries, including law, marketing, consulting, and IT services, by industry, services offered, location, and size. More granular skills like find-marketing-agency, find-seo-agency, find-design-agency, find-software-developer, find-web-developer, find-ai-consultancy, find-law-firm, find-cpa-firm, find-management-consultant, and find-engineering-firm offer specialized search capabilities within their respective domains. For example, find-ai-consultancy can locate firms specializing in AI/ML development, MLOps, and generative AI applications like RAG and chatbots. The skills are designed to be compatible with over 19 AI agents, including Claude Code, Codex, Cursor, GitHub Copilot, and Gemini, supporting the Agent Skills format.

Installation can be done by adding the entire nostrband/servicegraph package or specific skills using npx skills add. An API key is required for accessing data, with 2,000 free credits provided upon signup. The API key is managed securely and not directly exposed to the AI model's context.

Why it matters for builders

ServiceGraph empowers AI builders by providing a standardized interface to a rich set of business intelligence data. This integration allows developers to enhance their AI agents with capabilities for sophisticated data retrieval and analysis, moving beyond general knowledge to specific, actionable business insights. The availability of specialized skills for various professional services and business categories simplifies the process of building agents that can perform complex research and recommendation tasks, such as identifying suitable agencies for a project or finding companies with specific expertise.

Practical impact

Developers can integrate ServiceGraph skills into their AI applications to automate tasks such as market research, lead generation, and vendor selection. For instance, an agent could be tasked with finding marketing agencies specializing in B2B SaaS or identifying AI consultancies capable of building RAG systems. The structured nature of the data and the defined filter DSL simplify prompt engineering and improve the reliability of AI-driven business intelligence. Builders can explore the available skills on the ServiceGraph GitHub repository and integrate them into their preferred AI agent frameworks. The free credits offered upon signup allow for immediate experimentation and testing of the skills' capabilities.

Caveats and source limits

The current dataset primarily focuses on US professional-services firms and business directories, with limited coverage for other regions or dataset types. While more datasets are listed as 'coming soon,' their availability and specific enrichment details are not yet defined. The source does not provide independent benchmark results for the performance or accuracy of the skills. Pricing beyond the initial free credits is not detailed in the provided excerpts, and specific details on the enrichment sources for each dataset are not fully elaborated. The project is presented as a GitHub repository signal, and details regarding its official launch or widespread adoption are not available.