Outdooriq Mcp

io.github.bch1212由 GitHub 账号 bch1212 发布v0.1.0更新于 Nov 2, 2026

Paid MCP for fishing/lake/stocking intelligence — 72k+ US lakes, 293k+ events.

已验证Streamable HTTP可网页运行位置与生活数据与分析

概览

AI 生成的概览

付费远程 MCP 服务器,为助手提供美国湖泊、鱼类放流记录、垂钓适宜度评分和实时天气数据,用于行程与垂钓推荐。

可以做什么
  • 按名称、州、县或面积搜索和浏览湖泊,并获取坐标、深度、鱼种、设施等详情
  • 按 GPS 坐标查找附近湖泊并按 0-100 垂钓适宜度评分排序,也可按州或鱼种列出高分湖泊
  • 按湖泊、鱼种、年份或月份查询放流历史与近期放流计划,并列出当前有放流的鱼种
  • 获取湖泊当前天气与 7 天预报,并生成自然语言的垂钓报告摘要
适用场景
适合构建或运行垂钓、行程规划、旅行管家或户外品牌类助手,需要推荐湖泊、围绕放流事件安排出行时间,或按需生成垂钓报告,而不必自行整合多个数据源。
运行要求
远程 streamable HTTP 端点,无需本地安装包。需要通过 X-API-Key 请求头提供 API 密钥。提供免费开发密钥,每天限 50 次调用;Pro 为每月 14 美元,按次计费为每次 0.01 美元。天气数据来自 Open-Meteo,无需密钥。
安装前请注意
这是付费服务:超出免费每日额度后,按月或按次都会产生费用。API 密钥是机密信息,通过 X-API-Key 请求头发送,不应外泄。查询内容(湖泊 ID、坐标、鱼种筛选)会发送到第三方端点。README 提到数据库不可用时会退回 SQLite 种子数据,数据量小得多,结果可能受限。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Outdooriq Mcp,将其添加到工作区。
  2. 连接 X-API-Key 凭据。它们会被私密存储,绝不会在公开页面上展示。
  3. 为需要使用其工具的对话启用该服务。

Web executable,通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。

交给智能体安装

把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体,或 SourceWeft 对话中。智能体会读取该服务器的安装指南,先向你展示发布者、端点和登录方式,经你同意后再把它添加到你的客户端。密钥不会写进任何命令。

阅读 https://sourceweft.com/mcp/io.github.bch1212%2Foutdooriq-mcp/install.md,并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前,先告诉我它的发布者、端点和登录方式,等我确认。不要把 API 密钥写进命令或我会提交的文件;问我密钥应从哪里获取。

查看智能体所依据的安装指南

其他 MCP 客户端

把它添加到你客户端的 mcpServers 配置中。

{
  "mcpServers": {
    "outdooriq": {
      "type": "http",
      "url": "https://mcp.castiq.net/mcp",
      "headers": {
        "X-API-Key": "<YOUR_API_KEY>"
      }
    }
  }
}

README

OutdoorIQ MCP — Outdoor Recreation Intelligence

Powered by 72,000+ US lakes and 293,000+ stocking events across 12 states.

OutdoorIQ MCP is a paid MCP server that exposes lake conditions, fish-stocking records, fishing-favorability scoring, and live weather to AI agents. It is built on the CastIQ dataset (the same data that powers fishing-seo.pages.dev and the CastIQ catalog APIs).

If you're building a trip-planning assistant, a fishing app, a travel concierge, or an outdoor-brand agent that needs to recommend lakes, time visits to recent stocking events, or pull a fishing report on demand, this is the data layer you wire up.

For consumer-facing AI products, OutdoorIQ replaces a ten-source ETL with one authenticated MCP endpoint — and bills predictably so you don't get a surprise S3 invoice the first weekend traffic spikes.


Pricing

TierPriceLimits
Free$0outdooriq-dev-key-001, 50 calls/day
Pro$14/moUnlimited
Pay-as-you-go$0.01/callNo monthly minimum

Listing & checkout: https://mcpize.com/outdooriq-mcp


Install in Claude

Live URL: https://mcp.castiq.net/mcp (Railway fallback: https://web-production-9b8950.up.railway.app/mcp)

The fastest path uses mcp-remote as a stdio→HTTP bridge:

bash
claude mcp add outdooriq-mcp -- npx -y mcp-remote \  https://mcp.castiq.net/mcp \  --header "X-API-Key:outdooriq-dev-key-001"

Or configure manually:

json
{  "mcpServers": {    "outdooriq-mcp": {      "url": "https://mcp.castiq.net/mcp",      "headers": { "X-API-Key": "outdooriq-dev-key-001" }    }  }}

Anthropic MCP Registry entry: io.github.bch1212/outdooriq-mcp — listed at https://registry.modelcontextprotocol.io.

Example agent prompt:

"Find top 5 trout lakes near Chicago for this weekend."

The agent calls get_nearby_lakes (lat/lng of Chicago, radius 200mi), filters for species: trout via get_top_lakes, then pulls get_fishing_report_summary for the top match.


Tool Reference

ToolArgsReturns
search_lakesname?, state?, county?, min_acres?, max_acres?, limit?List of matching lakes
get_lake_detailslake_idCoords, acreage, depth, species, facilities
get_stocking_datalake_id?, species?, year?, limit?Recent stocking events
get_fishing_scorelake_id0-100 score with bucket breakdown
get_nearby_lakeslat, lng, radius_miles?, min_score?, limit?Lakes near GPS, sorted by score
get_weather_for_lakelake_idCurrent + 7-day forecast (Open-Meteo)
get_top_lakesstate?, species?, limit?Highest-scoring lakes
get_stocking_schedulestate?, species?, month?, limit?Most-recent matching events as a planning prior
search_speciesstate?, season?Actively-stocked species
get_fishing_report_summarylake_idNatural-language report

Architecture

client (Claude / agent)        │  HTTP POST /mcp  (JSON-RPC 2.0)        ▼   FastAPI app (server.py)        │  X-API-Key auth + per-day rate limiter        ▼  Tool registry (10 tools)        │        ▼  db.connection.py  ──────┐        │ Postgres mode  │  → CastIQ Postgres (72k lakes, 293k stockings)        │ SQLite fallback│  → bundled seed (100+ lakes, ~150 stockings)        ▼  tools.* (lakes, stocking, scoring, weather, reports)        │        ▼  Open-Meteo (no key)

Postgres vs SQLite mode

The server picks a backend at startup:

  1. If DATABASE_URL is set, it tries asyncpg.create_pool. On success → logs [OutdoorIQ] Running in Postgres mode.
  2. If the pool fails (timeout, bad creds, DB down) or DATABASE_URL is unset, the server seeds an in-memory SQLite DB with 100+ real lakes and logs [OutdoorIQ] Running in SQLite fallback mode.

The server always starts, regardless of Postgres availability.

CapabilityPostgres modeSQLite fallback
Lake catalog72,669 (12 states)~106 (WI, MN, IL, IA, MO)
Stocking events293,821 historical~150 templated, recency-tuned
Species coverageAll states/species in CastIQCurated subset (walleye, bass, musky, trout, crappie, perch, pike, panfish, salmon, lake_trout, sauger, white_bass, sturgeon, catfish, bluegill, smallmouth_bass)
Year-over-year analysisYes — multi-year stockingsLimited — events are anchored relative to "now"
Stocking-schedule toolReal historical patternsApproximated from seed
Weather, scoring, reportsIdenticalIdentical

Local development

bash
git clone <this repo>cd mcp-outdoorspython3 -m venv .venv && source .venv/bin/activatepip install -r requirements.txt
# Run with SQLite fallback (no DATABASE_URL needed)python -m run
# OR run against the CastIQ Postgresexport DATABASE_URL=postgresql://vikinetic:vikinetic_dev@localhost:5444/vikineticpython -m run

Then:

bash
curl -s http://localhost:8080/healthcurl -s -X POST http://localhost:8080/mcp \  -H "X-API-Key: outdooriq-dev-key-001" \  -H "Content-Type: application/json" \  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Tests

bash
pytest -v

The test suite covers both the SQLite fallback path and the Postgres dispatch path (via a mocked asyncpg pool), plus auth, rate limits, all 10 tools, JSON-RPC initialize / list / call, and the scoring algorithm's bucket math.


Deploy to Railway

A deploy.sh script is included at the repo root. Run it on your Mac (the Cowork sandbox can't reach Railway/Stripe/Cloudflare APIs):

bash
./deploy.sh

It expects RAILWAY_API_TOKEN (or RAILWAY_TOKEN exported as RAILWAY_API_TOKEN), points the project at nixpacks.toml, and sets the DATABASE_URL env var if you've also provisioned the CastIQ Postgres on Railway.

Railway gotcha (already handled): Railway exec's startCommand without a shell, so $PORT doesn't expand. We use python -m run and read PORT from os.environ inside run.py.


MCPize listing copy

OutdoorIQ MCP — the data layer for outdoor-rec AI agents. 72,000+ US lakes, 293,000+ fish-stocking events, and live weather behind one authenticated endpoint. Search lakes by name, state, or acreage; pull stocking history filtered by species and month; compute a 0-100 fishing-favorability score that bakes in recency, species diversity, lake size, and current conditions. One JSON-RPC call replaces a multi-source ETL.

Built for fishing apps, trip-planning assistants, travel concierges, and outdoor-brand agents. $14/mo Pro for unlimited use, or pay $0.01 per call. Free dev tier (50 calls/day) lets you ship a prototype before opening your wallet. Install with claude mcp add outdooriq-mcp --url https://mcp.castiq.net/mcp.


License

MIT. See LICENSE.

来源:README.md,提交 2cc3026

工具

0
该服务器需要 API 密钥等凭据,因此 SourceWeft 无法在没有凭据的情况下读取其工具,也未收录。

版本历史

1
  1. v0.1.0最新Sep 16, 2026

评分

已满足适用检查项的 85%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通必需25 分
  • 已声明认证方式必需10 分
  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分