# FRED Tags & Sources MCP MCP > FRED Tags & Sources — Data Discovery helps your AI agent pinpoint exact economic data series across all 107 official sources. Instead of guessing, you search by category, geography, or frequency tags—like 'gdp' or 'monthly'—to find the precise dataset ID needed for analysis. ## Overview - **Category:** brain-trust - **Price:** Free - **Tags:** data-tagging, metadata, search-filtering, economic-data, data-discovery, classification ## Description Need to pull a specific piece of macroeconomics data but don't know its series ID? This MCP is your discovery layer. It lets your agent drill down into FRED’s entire catalog, moving you past simple keyword searches and straight to metadata precision. You can narrow results by combining tags—say, finding every 'gdp' measure that was reported 'monthly' for the 'usa'. You also get a full list of all organizations contributing data, from the BLS to the Census Bureau, so your agent knows exactly where to look. When you connect this MCP via Vinkius, your AI client can map these complex relationships and retrieve metadata faster than manual browsing ever could. ## Tools ### search_tags Allows you to browse or search for metadata tags related to geography, topic, source, or frequency. ### get_series_by_tags Finds specific FRED series by combining multiple tags, letting you narrow results using required or excluded criteria. ### list_sources Provides a complete list of every organization that contributes data to the FRED database. ## Prompt Examples **Prompt:** ``` Find all monthly U.S. GDP-related series ``` **Response:** ``` 🏷 **Tags: usa + gdp + monthly** Found 47 matching series: 1. MRTSSM44X72USS — Retail Sales 2. TOTALSA — Vehicle Sales 3. INDPRO — Industrial Production 4. TSIFRGHT — Transportation Index ... All are monthly U.S. GDP-component indicators. ``` **Prompt:** ``` List all data sources that contribute to FRED ``` **Response:** ``` 🏛 **FRED Data Sources (107 total)** 1. Bureau of Labor Statistics (BLS) 2. Bureau of Economic Analysis (BEA) 3. Board of Governors of the Federal Reserve 4. U.S. Census Bureau 5. U.S. Department of the Treasury 6. International Monetary Fund (IMF) 7. World Bank 8. OECD ... 107 official sources spanning U.S. and international organizations. ``` **Prompt:** ``` What tags are most popular on FRED? ``` **Response:** ``` 🏷 **Top FRED Tags by Series Count** 1. `usa` — 245K series 2. `monthly` — 178K series 3. `annual` — 156K series 4. `nsa` (not seasonally adjusted) — 142K 5. `county` — 98K series 6. `quarterly` — 67K series 7. `sa` (seasonally adjusted) — 54K 8. `gdp` — 12K series ``` ## Capabilities ### Find series by multiple tags The MCP retrieves a list of data series that match specific combinations of tags like geography or topic. ### Browse all contributing organizations It lists every organization, such as the BEA or Treasury, that supplies data to FRED. ### Search by metadata type The MCP allows you to search across tags for location (europe), subject matter (inflation), or report frequency (quarterly). ## Use Cases ### A financial analyst needs to compare inflation rates. The agent uses `search_tags` to confirm the tag 'inflation'. It then calls `get_series_by_tags` using that tag combined with 'usa' and 'monthly' tags, instantly pulling a list of comparable series IDs for a report. ### A data engineer needs to map all possible sources. The agent runs `list_sources` first. This gives the engineer a comprehensive inventory of every organization feeding FRED, which is essential for mapping out potential API dependencies in an ETL pipeline. ### A researcher wants to find GDP data across continents. The agent uses `search_tags` to confirm 'gdp' as a topic tag. It then calls `get_series_by_tags`, combining 'europe' and 'gdp', letting them compare multiple international datasets in one query. ### A developer needs to understand data provenance. The agent runs `list_sources` and then uses the output of that list to inform which specific tags must be used with `get_series_by_tags`, ensuring the resulting dataset is trustworthy. ## Benefits - Stop guessing which data source to use. `list_sources` gives your agent a full directory of all 107 contributing organizations, so you know exactly where the number comes from. - Pinpoint specific datasets instantly. By using `get_series_by_tags`, you can combine tags like 'usa', 'gdp', and 'quarterly' to find only the exact series ID needed for your analysis. - Filter complex metadata accurately. The `search_tags` tool lets you browse or filter by any dimension—be it a geographic area, an economic topic, or reporting frequency. - Saves hours of manual searching. Instead of clicking through multiple tabs on the FRED site, your agent handles the multi-tag logic in one go. - Works with massive data sets. This MCP is built to handle large catalogs, meaning you don't hit a wall when looking for rare or niche economic indicators. ## How It Works The bottom line is your agent uses these tools sequentially—search tags first, then get the specific ID—to build an accurate data pointer. 1. Start by using the `search_tags` tool to browse or filter metadata tags, identifying key criteria like 'usa' and 'gdp'. 2. Next, pass those identified tags into `get_series_by_tags` to narrow the scope and find series that match all specified conditions. 3. Finally, run `list_sources` if you need a full list of contributing organizations for context on where the data originates. ## Frequently Asked Questions **How do I use get_series_by_tags to limit my search?** You must provide specific tags. For example, passing a combination like 'usa;gdp+monthly' ensures the agent only returns series matching all three criteria. **Does list_sources tell me if the data is complete?** The `list_sources` tool just lists the contributing organizations (like BLS or Census Bureau). It tells you *who* provides the data, not whether that specific dataset has gaps. **What tags can I use with search_tags?** You can browse by geography (usa), topic (inflation), source (bea), or frequency (quarterly). This makes it easy to define your search boundaries. **Can get_series_by_tags handle multiple countries?** Yes. You combine the country codes as tags. For example, you'd tag both 'usa' and 'europe' along with the topic tag to compare them. **What happens when I run `get_series_by_tags` with conflicting or non-existent tags?** The tool handles this by returning a specific error code and an empty list of results. This allows your AI client to easily detect invalid inputs and adjust the query, preventing pipeline failures. **How do I handle rate limits when frequently using `search_tags`?** If you hit a limit, wait a short period (a few seconds) before retrying. Your agent should implement an exponential backoff strategy to ensure reliable data retrieval without overloading the API. **What structured output does `list_sources` provide?** It outputs a clear, machine-readable list of all contributing organizations. Each result includes the source name and often a brief description, making it easy for your agent to build a definitive roster. **Does `get_series_by_tags` require an authentication key?** Yes, you'll need proper credentials configured in your MCP environment. The connection must be authenticated before the tool can successfully query and pull data from FRED. **What is the difference between tags and categories?** Categories are a strict hierarchy (one parent, many children). Tags are flat labels — each series can have many tags across dimensions: geography (usa), frequency (monthly), topic (gdp), source (bls). Combine tags for powerful cross-dimensional filtering.