{
  "markdown": "# fred-mcp\n\n<!-- mcp-name: io.github.mcpwright/fred-mcp -->\n\n**FRED® economic data inside your AI agent — including what the numbers said\n*before* the revisions.** An MCP server for the Federal Reserve Bank of\nSt. Louis's FRED® API: search and read 800,000+ U.S. and international time\nseries, follow the release calendar, and — the distinctive part — query\n**ALFRED® vintage data**: any series exactly as it was known on a past date,\nand any single data point's full revision history from its initial print to\ntoday's value.\n\nRead-only, built on public endpoints, powered by **your own free FRED API\nkey**. Live API calls with a volatility-aware cache (vintage reads are\nimmutable and cache long; current data stays fresh) — no bulk mirroring.\n\n> This product uses the FRED® API but is not endorsed or certified by the\n> Federal Reserve Bank of St. Louis. By using this server you agree to the\n> [FRED® API Terms of Use](https://fred.stlouisfed.org/docs/api/terms_of_use.html).\n\n**Status:** v0.1.0 — live on [PyPI](https://pypi.org/project/mcpwright-fred/)\n(`uvx mcpwright-fred`) and the official MCP Registry\n(`io.github.mcpwright/fred-mcp`). 9 tools, CI-gated, built on\n[mcpwright-core](https://github.com/mcpwright/mcpwright-core).\n\n## Why vintage data matters\n\nEconomic numbers get revised — sometimes dramatically. The *advance* estimate\nof Q4-2008 real GDP told policymakers the economy was shrinking far more\nslowly than the number that stands today. If an agent reasons about what\ndecision-makers knew *at the time* using *current* data, it is quietly wrong.\n`get_series_as_of` and `get_revision_history` make as-published data a\nfirst-class question.\n\n## Tools\n\n| Tool | What it does |\n| --- | --- |\n| `search_series` | Find series IDs by free text, best matches first |\n| `get_series` | One series' full metadata + verbatim notes |\n| `get_observations` | A date window of values; transforms (`pct_change_yoy`, …) + downsampling |\n| `get_latest` | The most recent value + the next scheduled release date |\n| `compare_series` | 2–5 series aligned on one date axis |\n| `get_release_calendar` | Upcoming data releases (\"when's the next jobs report?\") |\n| `get_series_as_of` ⭐ | A series exactly as known on a past date (ALFRED vintage) |\n| `get_revision_history` ⭐ | One data point's life: initial print → every revision → today |\n| `get_vintage_dates` ⭐ | When a series was released/revised; bounds the vintage tools |\n\nAll tools are read-only and annotated as such.\n\n## What can you ask?\n\nEvery question below was run live against the real FRED API through these\ntools (answers as of June 2026 — ask again and they'll be current).\n\n**Everyday questions:**\n\n1. *\"What's the unemployment rate right now, and when's the next jobs\n   report?\"* — `get_latest(UNRATE)` → 4.3% (May 2026), next release 2026-07-02.\n2. *\"What's the 10-year Treasury yield today?\"* — `get_latest(DGS10)` → 4.53%\n   (it correctly skips the `.` rows daily series publish on holidays).\n3. *\"How has inflation trended over the past year?\"* —\n   `get_observations(CPIAUCSL, transform=pct_change_yoy)` → 2.4% → 3.3% →\n   3.8% → 4.2% over Feb–May 2026.\n4. *\"Find me data on median household income.\"* — `search_series` →\n   `MEHOINUSA672N` (real) and `MEHOINUSA646N` (nominal), annual.\n5. *\"Is the labor market loosening? Compare unemployment and job openings\n   since 2024.\"* — `compare_series([UNRATE, JTSJOL])` → one aligned table.\n6. *\"What economic data comes out this week?\"* — `get_release_calendar(7)` →\n   241 release dates, CPI included.\n7. *\"What was real GDP growth each quarter of 2025?\"* —\n   `get_observations(A191RL1Q225SBEA)` → −0.6, +3.8, +4.4, +0.5%.\n8. *\"Give me CPI as annual averages for the 2020s.\"* —\n   `get_observations(CPIAUCSL, frequency=annual)` → one row per year.\n\n**Questions only the vintage tools can answer:**\n\n9. *\"How bad did Q4-2008 GDP look to policymakers in early 2009, versus what\n   we know now?\"* — `get_series_as_of(GDPC1, \"2009-02-15\")` → a −3.8%\n   annualized decline as known then (in chained-2000 dollars — the metadata\n   is vintage-pinned too) versus −8.5% in today's data. Policy ran on the\n   first number.\n10. *\"Was the May 2020 COVID unemployment rate ever revised?\"* —\n    `get_revision_history(UNRATE, \"2020-05-01\")` → first published as 13.3\n    (2020-06-05), currently 13.2, with each revision dated.\n11. *\"What mortgage rate did the Fed see going into the March 2022 liftoff\n    meeting?\"* — `get_series_as_of(MORTGAGE30US, \"2022-03-16\")` → 3.85% as\n    known then; 6.48% today.\n12. *\"How often does GDP actually get revised?\"* — `get_vintage_dates(GDPC1)`\n    → 415 vintages since 1991-12-04.\n\n## Install\n\nYou need a **free FRED API key** (takes seconds):\n<https://fred.stlouisfed.org/docs/api/api_key_request.html>. Keys are personal\nunder the FRED terms — bring your own.\n\n### `uvx` (any MCP client)\n\n```json\n{\n  \"mcpServers\": {\n    \"fred\": {\n      \"command\": \"uvx\",\n      \"args\": [\"mcpwright-fred\"],\n      \"env\": { \"FRED_API_KEY\": \"your-key-here\" }\n    }\n  }\n}\n```\n\n### Claude Code\n\n```bash\nclaude mcp add fred -e FRED_API_KEY=your-key-here -- uvx mcpwright-fred\n```\n\n### Claude Desktop\n\nUse the one-click `.mcpb` extension from the\n[latest release](https://github.com/mcpwright/fred-mcp/releases) (it prompts\nfor your API key), or add the `uvx` JSON above to your\n`claude_desktop_config.json`.\n\n### OpenAI Agents SDK / other clients\n\nAny MCP-capable client works — point it at `uvx mcpwright-fred` over stdio\nwith `FRED_API_KEY` in the environment.\n\n## Notes\n\n- **Revisions:** current values often differ from what was originally\n  published; for \"what was known at the time\", use the vintage tools.\n- **Copyrighted series:** some FRED series are owned by third parties; their\n  notes carry the owner's terms (surfaced verbatim, flagged via\n  `copyrighted`). You are responsible for complying with them beyond personal\n  use.\n- **Caching & rate limits:** responses are cached in-memory (vintage reads\n  long, live reads short) and requests are throttled well under FRED's rate\n  limit. Set `FRED_MCP_CACHE=0` to disable caching, `FRED_MCP_USER_AGENT` to\n  identify your own deployment.\n\n## Develop\n\n```bash\ngit clone https://github.com/mcpwright/fred-mcp && cd fred-mcp\nuv sync\nuv run pytest                                  # tests\nuv run mypy                                    # types\nuv run ruff check src/ && uv run ruff format --check src/   # lint + format\nuv run pre-commit run --all-files              # everything, like CI\n```\n\nDev loop: feature branch → PR → `Code Quality & Tests` green → squash-merge.\n\n## Roadmap\n\n- [x] v1 tool surface (9 tools incl. the ALFRED vintage lane)\n- [x] Publish to PyPI (`mcpwright-fred`) + the MCP Registry\n- [x] Site page at [mcpwright.com/fred](https://mcpwright.com/fred)\n- [x] `.mcpb` one-click Claude Desktop extension ([v0.1.0 release](https://github.com/mcpwright/fred-mcp/releases/tag/v0.1.0))\n- [ ] GeoFRED / maps (pairs with [census-mcp](https://github.com/mcpwright/census-mcp))\n\n## Questions & feedback\n\n[Discussions](https://github.com/mcpwright/fred-mcp/discussions) for questions\nand ideas · [Issues](https://github.com/mcpwright/fred-mcp/issues) for bugs.\n\n---\n\nPart of **[mcpwright](https://mcpwright.com)** — polished MCP servers for\nauthoritative public data · built by [Devender Gollapally](https://devender.me).\nFRED® and ALFRED® are registered trademarks of the Federal Reserve Bank of\nSt. Louis. This project is not affiliated with, endorsed, or certified by the\nFederal Reserve Bank of St. Louis.\n",
  "bytes": 7504,
  "sha": "d561235d2bd6f68a0e933d99b9fd4a2a22ec2e6bc73e194aa8bdfb4978a46d4a",
  "repo_slug": "mcpwright/fred-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_mcpwright_fred_mcp_aa69a57c/readme"
}