# SignalForge is your research intelligence MCP. AI Agent Connect > SignalForge fuses seven public, keyless data sources into a single radar for your AI agent. It pulls from OpenAlex, Crossref, OpenAIRE, Europe PMC, arXiv, CORDIS, and an experimental USPTO patent layer to show you exactly where a field is moving. You get publication momentum, funding data, and patent activity without needing a single API key. ## Overview - **Category:** research - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_tT0gbLZM13D9lVffInnHXYgQHX7e6QF8eykgddKR/ai-agent-connect - **Tags:** openalex, crossref, openaire, europe-pmc, arxiv, cordis, uspto, patents, funding, preprints, trends, keyless ## Description You can stop chasing old news and start watching the data move in real time. SignalForge acts as a high-speed sensor for your AI client, pulling from a massive web of scholarly and patent data to identify momentum. Instead of searching through individual databases, you use this MCP to fuse disparate signals like arXiv preprints, EU funding projects, and patent filings into a unified view. It is built to be resilient; if one source like the experimental USPTO leg is down, the rest of your scan still works. You can track the growth of specific topics through ASCII-rendered momentum bars or dig into who is actually funding the research through Crossref metadata. It is a tool for seeing the inflection points in science and technology before they become obvious trends. ## Tools ### openalex_topics This tool identifies specific topic and concept clusters within OpenAlex. It helps your agent monitor discrete channels by providing corpus size and citation counts for each cluster. ### cordis_project This tool pulls a complete fact sheet for specific EU innovation projects. It provides the grant ID, coordinator, EU contribution, and objectives directly from CORDIS. ### crossref_search This tool queries Crossref for publication metadata and funding details. Use it to identify which specific funders are backing research in a particular field. ### europepmc_search This tool searches the Europe PMC database for biomedical articles and preprints. It allows your agent to rank results by citation impact or recency. ### openaire_search This tool accesses OpenAIRE for EU research intelligence, including publications and funded projects. It includes flags for open-access content. ### openalex_search This tool searches the OpenAlex corpus of over 250 million works. It returns title, author, and citation data, allowing your agent to gauge field size or freshness. ### openalex_trend This tool calculates publication momentum using per-year counts. It returns ASCII bars and growth percentages to show how fast a topic is accelerating. ### patent_signals This tool provides a best-effort look at patent filing activity. It uses an experimental USPTO leg to surface commercial intent ahead of product launches. ### signal_scan This tool runs a simultaneous scan across seven different public sources. It returns the status of each source and the top signals found in each. ### arxiv_emerging This tool pulls the newest submissions from arXiv. It provides abstracts and PDF links for research that is only days old. ## Prompt Examples **Prompt:** ``` Any early signals in solid-state batteries? ``` **Response:** ``` The signal_scan shows the full radar: freshest OpenAlex works, patent filing activity, latest arXiv preprints, biomedical crossovers in Europe PMC, EU projects, and Crossref funders. Each source is marked with its current status so you know the data reliability. ``` **Prompt:** ``` Which fields are accelerating right now in AI safety? ``` **Response:** ``` Running openalex_trend on topics like 'interpretability' or 'mechanistic alignment' shows per-year publication counts as ASCII bars and growth percentages to rank momentum. You can then use openalex_topics to break those winners into specific sub-topics to monitor. ``` **Prompt:** ``` Who in the EU is funded to work on green hydrogen? ``` **Response:** ``` Using openaire_search with the term 'green hydrogen' lists the relevant EU-funded projects and their summaries. You can then use cordis_project to pull the full fact sheet for a specific project, including the coordinator and EU contribution. ``` ## Capabilities ### Multi-source fusion Your agent pulls from seven different databases in a single command. ### Momentum tracking The agent calculates growth percentages and visualizes publication trends. ### Funding intelligence Your agent identifies specific organizations and agencies investing in new fields. ### Real-time preprints The agent accesses the freshest research from arXiv and Europe PMC. ### Commercial intent detection Your agent monitors patent filing activity to spot upcoming technology shifts. ## Use Cases ### Spotting Tech Inflections An investor uses openalex_trend to find a niche field with a 200% annual growth rate. ### Competitive Intelligence An R&D lead uses patent_signals to see if a competitor is ramping up patent filings in a specific area. ### Funding Analysis A researcher uses crossref_search to see which government agencies are currently funding green energy. ### Literature Reviews A scientist uses arxiv_emerging to find the most recent papers published in the last few days. ## Benefits - No API keys required for any of the seven data sources. - One call to signal_scan aggregates data from all providers. - Resilient architecture ensures one failing source doesn't break the whole scan. - Visual ASCII bars make publication growth easy to read in a chat interface. ## How It Works Connecting to SignalForge gives your AI client immediate access to global research data. 1. Connect the SignalForge MCP to your AI client via Vinkius. 2. Ask your agent to scan for a specific keyword or technology. 3. The agent calls the necessary tools to pull data from OpenAlex, arXiv, and other sources. 4. The agent presents a unified view of trends, funding, and recent papers. ## Frequently Asked Questions **Do I need an API key?** No. OpenAlex, Crossref, OpenAIRE, Europe PMC, arXiv and the CORDIS public pages are all keyless. The USPTO Patent Public Search leg is experimental and keyless only until 2026-11-07, when USPTO makes an account mandatory — that leg degrades to a structured unavailable note, and no keyless patent fallback exists today (OpenAlex public type:patent index is empty, probed 2026-09-30), so the patent layer is best-effort. **What happens when one source is down?** Every source is an independent leg. signal_scan runs all legs in parallel and contains each failure inside its own leg — the report marks that leg 'unavailable' with the reason and returns the rest. A source under rate limit is marked 'degraded' with a retry hint instead of failing the call. **Why is The Lens not a source?** www.thelens.org was observed parked (a GoDaddy for-sale redirect) in September 2026, so the platform was deliberately excluded. The patent layer The Lens offered is not covered by any keyless source today: OpenAlex public type:patent index is empty (0 records, probed 2026-09-30) and the USPTO leg is experimental, keyless only until 2026-11-07. **How do I see EU-funded projects on a keyword?** CORDIS's own search backend is not anonymously reachable, so keyword discovery of EU projects flows through OpenAIRE projects (which aggregates CORDIS and EC grants). Pick a project code from openaire_search (resource='projects') and open its full fact sheet with cordis_project. **Is there rate limiting?** OpenAlex rate-limits anonymous bursts (429 with retry-after). SignalForge serialises all OpenAlex calls in one process and spaces per-year trend reads, so a scan or trend stays inside the anonymous window; when a limit is hit the leg reports 'degraded' with a retry hint instead of erroring.