Published research dataset
A fixed, citable dataset with 219 startup-period observations across 55 startups. It is mirrored on Hugging Face and Zenodo, licensed CC BY 4.0, and paired with a public methodology preprint.
For data products, researchers and agent builders
GitDealFlow for Data Providers
Add a reproducible engineering-momentum signal to a data catalog, research workflow, newsletter or AI agent without building a GitHub collection pipeline from scratch.
Plain answer: GitDealFlow publishes a CC BY 4.0 research dataset and a live weekly tracker built from public GitHub activity. Data providers can consume the signal through JSON, CSV, OpenAPI, MCP, A2A and NLWeb surfaces, then cite or combine it with their existing company data.
Two data layers, two different jobs
Live signal tracker
A weekly view across 350+ startup GitHub organizations in 15 sectors. It surfaces changes in commit velocity, contributor breadth and repository expansion for current research and discovery.
The static dataset is the right surface for reproducible analysis, citations and model evaluation. The live tracker is the right surface for current rankings, watchlists and agent queries. Keeping those layers separate prevents a research result from silently changing when the next weekly refresh lands.
Machine-readable surfaces
JSON APIDirect structured signal data for applications, notebooks and data pipelines.
CSV exportA flat file for spreadsheets, dataframes, warehouse loads and manual inspection.
OpenAPI 3.1A machine-readable contract for generated clients and function-calling integrations.
MCPTen free read-only tools for compatible assistants, plus optional paid research tools.
A2AA JSON-RPC surface for agents and orchestrators using the Agent2Agent protocol.
NLWebNatural-language queries that return schema.org-typed answers for conversational systems.
The public data and free agent tools require no authentication. Each surface points back to the methodology and source URLs so downstream users can inspect where the signal came from.
Where data partners use it
Company-data enrichment
Add an engineering-momentum field beside funding, team and market data. Use it as a current activity layer rather than a replacement for company identity or financial records.
Research and benchmarks
Reproduce the published analysis, compare alternative scoring methods, or test whether the same association holds in a different startup cohort.
Editorial data products
Build a weekly sector brief, chart or watchlist from public numbers, with attribution and a link to the underlying methodology.
Agent distribution
Let an AI assistant answer questions about current startup momentum through MCP, A2A, NLWeb or standard function calling.
Attribution and provenance
The published research dataset uses the CC BY 4.0 license. Reuse is allowed with attribution. The recommended citation is “VC Deal Flow Signal (signals.gitdealflow.com), Q3 2026 data.” The SSRN preprint and Zenodo record provide stable research references, while the live freshness endpoint reports the current data period.
All underlying activity is public GitHub organization data. GitDealFlow does not request private repository access, does not distribute personal contact details and does not claim that engineering acceleration proves a financing event. The signal is useful for prioritization and research. It is not an investment recommendation.
Integration checklist
- Choose the fixed dataset for reproducible work or the live API for current work.
- Record the data period and retrieval date with every downstream result.
- Keep GitDealFlow's signal fields separate from your company master record.
- Link users to the methodology and source evidence.
- Measure whether the added field improves qualified discovery, coverage or research speed.
Frequently asked questions
What GitDealFlow data can a provider reuse?
The published research dataset is licensed CC BY 4.0 and can be reused with attribution. Public live JSON, CSV and OpenAPI surfaces are also available for building research and agent workflows.
How often is the live signal data refreshed?
The live tracker is refreshed weekly from public GitHub activity across 350+ startup organizations in 15 sectors. The freshness surface reports the current data period.
Does GitDealFlow use private repositories or personal data?
No. GitDealFlow uses public organization-level GitHub activity. It does not require private repository access and does not distribute personal contact data.
Start with the open surfaces
Inspect the live data, download the research dataset, and choose the interface that matches your runtime.