Clairwave

io.github.clairwave由 GitHub 账号 clairwave 发布v0.4.0更新于 Nov 2, 2026

Ocean acoustics: propagation models, bathymetry, sound speed, vessel noise, live AIS. Open.

已验证Streamable HTTP可网页运行位置与生活数据与分析AI 与机器学习

概览

AI 生成的概览

开放、免认证的远程 MCP 服务器,提供海洋声学能力:水深、声速剖面、传播与声呐模型,以及实时 AIS 船舶数据。

可以做什么
  • 把地名解析为水域坐标,支持向海侧吸附与离岸偏移
  • 获取单点或沿方位断面的海底深度,以及季节性声速剖面与海底参数
  • 运行传播损失模拟(RAM 抛物方程、三维 Bellhop 体积)并估算声呐探测距离
  • 按船名或 MMSI 搜索实时 AIS 船舶、列出某点附近船舶,并获取单船航迹、静态资料与三维模型
  • 获取船舶辐射噪声源级,以及鱼场、水听器等固定站点的功率合成接收级
  • 获取带署名的船舶照片,并查看各工具背后的模型与数据说明
适用场景
适合需要定量分析水下声音、声呐覆盖、航运噪声或海底条件,而不是凭猜测作答的场景,也适合关注某位置附近实时船舶动态。结果为基于气候态平均的筛查性估算,不能替代现场实测。
运行要求
远程 Streamable HTTP 端点 clairwave.com/mcp,无需账号、API 密钥或请求头。若本地运行需 Python 及 mcp、httpx 包,并设置 CLAIRWAVE_API、CLAIRWAVE_FLEET、CLAIRWAVE_SITE、MCP_PORT 环境变量。需要网络访问。
安装前请注意
计算类调用共享约每分钟 20 次的平台总额度,频繁使用可能被限流。模拟基于月度气候态声速与全球水深数据,属筛查性估算而非实测结果。船舶照片来自 Wikimedia Commons,使用时需保留署名。结果会以 run id 形式存储在平台上,并附带公开的打开链接。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Clairwave,将其添加到工作区。
  2. 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.clairwave%2Fclairwave/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": {
    "clairwave": {
      "type": "http",
      "url": "https://www.clairwave.com/mcp"
    }
  }
}

README

clairwave-mcp

An open MCP server that gives AI assistants physically grounded ocean acoustics.

Clairwave runs validated propagation models (Bellhop, RAM/parabolic equation) on global bathymetry and seasonal sound-speed profiles, tracks live AIS vessels, and serves 3D hull models for them (shipshape). This server exposes that to Claude, ChatGPT, Gemini and any other MCP client — so an assistant reasoning about the ocean can run the physics instead of guessing.

Every result carries provenance (model, data source, run_id) and an open_url that opens the exact result in the platform. Simulation results include the bathymetry, sound-speed profile and bottom parameters that were used, so a researcher can replicate the run in MATLAB, Python or anything else.

Endpoint (no auth, no key): https://www.clairwave.com/mcp — Streamable HTTP.

Connect

  • Claude Code: claude mcp add --transport http clairwave https://www.clairwave.com/mcp
  • Claude.ai / Claude Desktop: Settings → Connectors → Add custom connector → the URL above
  • ChatGPT: Plugins → search Clairwave → + (a listed plugin: no URL, no developer mode)
  • Any MCP client: point it at the URL; the server is stateless and JSON-response capable

Tools

ToolWhat it does
get_bathymetryDepth at a point, or a transect profile along a bearing
get_sound_speed_profileSeasonal c(z) for a month + seabed parameters (cp, cs, density, attenuation, sediment)
run_transmission_lossRAM parabolic-equation TL along a bearing; bathymetry/SSP/seabed fetched automatically; replication bundle included
estimate_detection_rangeSonar equation on a RAM run: continuous and furthest detection range, signal excess vs range
run_bellhop_volume3D Bellhop TL volume stored under a run id (uint8 cube + JSON sidecar links)
vessel_source_levelShip radiated noise: broadband + third-octave spectrum + mechanism breakdown
search_vessels / vessels_nearLive AIS by name/MMSI, or within a radius of a point
get_vesselLive position/track, particulars, and the 3D model (GLB, bow=+Z) with platform links
get_vessel_photoWikimedia Commons photo with attribution
resolve_placePlace name (port, strait, sea, 'off Halifax') → water coordinates; gazetteer + OpenStreetMap, snapped seaward off land
habitat_received_levelPower-summed vessel noise at a fixed site (fish farm, reef, hydrophone): live snapshot or 10-minute history series; top contributors
aboutModels, data sources, limits

Typical latency against the live platform: bathymetry 0.5 s, SSP 6 s first time per 0.1° cell then cached, RAM transmission loss 1–3 s, detection range 1–3 s.

Run locally

bash
pip install "mcp[cli]<2" httpxpython server.py            # streamable HTTP on :8890 (/mcp)python server.py --stdio    # stdio for local clientspython tests/smoke_client.py

Environment: CLAIRWAVE_API, CLAIRWAVE_FLEET, CLAIRWAVE_SITE, MCP_PORT.

Where to find it

  • Official MCP Registry: io.github.clairwave/clairwave (https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.clairwave/clairwave)
  • Claude: Settings > Connectors > Add custom connector, URL https://www.clairwave.com/mcp, no auth.
  • ChatGPT: listed in the Plugins directory as "Clairwave — Ocean acoustics & live ships"; search Clairwave and click +.
  • Grok (grok.com/connectors > New > Custom): paste the same URL.
  • xAI / OpenAI APIs: {"type": "mcp", "server_url": "https://www.clairwave.com/mcp", "server_label": "clairwave"}.

Place names

Every location tool takes either lat/lon or a place string. Names go through a maritime gazetteer first (ports resolve to their approaches, straits and seas to a representative water point; ~120 entries in gazetteer.py), then OpenStreetMap Nominatim. If the point is on land or shallower than 10 m it is walked seaward until it is deep enough, and the response's location block reports the original point, the snap distance and bearing, and the depth used. resolve_place exposes the same logic directly, with offshore_km to push a point further out.

Example prompts

  • "What is the sound speed profile 50 km west of Gibraltar in March, and where is the sonic layer depth?"
  • "How far could a 150 Hz, 170 dB source at 20 m depth be detected by a receiver at 100 m near 36N 5.5W, along bearing 090?"
  • "Show transmission loss versus range at 200 Hz out to 30 km north of Halifax in winter."
  • "What ships are within 15 km of the Strait of Hormuz right now, and how loud is the largest one?"
  • "Run a 3D Bellhop volume at 400 Hz around 49.2N 123.3W and give me the link to open it."

Limits and support

License

MIT. Data: AIS via the AISHub peer network (Clairwave contributes receivers); vessel photos CC-licensed with attribution; bathymetry and SSP sources cited in each response.

来源:README.md,提交 53e9f17

工具

13

resolve_place

Resolve place name

Turn a place name ('outside Halifax', 'Strait of Hormuz', 'Bergen, Norway') into water coordinates. Maritime gazetteer first (ports resolve to their approaches, straits/seas to a representative point), then OpenStreetMap. Points on land or shallower than min_depth_m are walked seaward along the gazetteer bearing (or the nearest direction) until deep enough; the original point and the snap are reported. offshore_km pushes the point further out along seaward_bearing_deg. All location tools also accept `place` directly, so a separate call is only needed to inspect.

输入参数place必填min_depth_moffshore_kmseaward_bearing_deg

get_bathymetry

Bathymetry

Seafloor depth from Clairwave-served global bathymetry (~450 m resolution). Location: lat/lon or `place` (name resolved to a water point, echoed in `location`). Without a bearing: depth at the point. With a bearing (compass degrees, 0 = north): a depth profile along that transect out to range_km.

输入参数latlonplacen_pointsrange_kmbearing_deg

get_sound_speed_profile

Sound speed profile

Seasonal sound-speed profile c(z) at a location (lat/lon or `place` name, e.g. "Halifax approaches") for a calendar month (1-12), from Clairwave-served seasonal temperature/salinity climatology, plus the seabed parameters (compressional/shear speed, density ratio, attenuation, sediment type) at the same point. Cached per 0.1 degree.

输入参数latlonmonth必填place

run_transmission_loss

Transmission loss (RAM PE)

Run a physically grounded transmission-loss simulation (RAM parabolic equation) from a source at (lat, lon or `place`, depth) along a compass bearing. Bathymetry, the seasonal sound-speed profile (`month`) and seabed parameters are fetched for the location automatically. Returns TL vs range at the receiver depths (default: source depth, 10 m, 100 m, mid-water), grid statistics, and the full replication bundle. Show `open_url` to the user as a clickable link.

输入参数latlonmonthplacerange_kmbearing_deg必填frequency_hz必填source_depth_m必填receiver_depths_m

estimate_detection_range

Detection range (sonar equation, passive or active)

Sonar detection range along a bearing from a location (lat/lon or `place`). Transmission loss (TL) is ALWAYS computed: RAM parabolic equation with the location's bathymetry, seasonal sound speed (`month`, default: current month) and seabed. Every other sonar-equation term is optional — anything not given gets a stated, typical default listed in `assumptions` with its source. Tell the user which values were assumed. Detection threshold DT is derived from processing gain (unless given): detection index d from `pd`/`pfa` (default 0.5 / 1e-4 → d ≈ 13.8), then passive (incoherent energy detector): DT = 5·log10(d / (B·T)), B = analysis band, T = `integration_time_s` (default 1 s); active (matched filter): DT = 10·log10(d / (2·B·τ)), B = `bandwidth_hz` (default 100 Hz LFM), τ = `pulse_length_s` (default 0.5 s). DT is in-band SNR, so it can be negative — that IS the processing gain. mode="passive": SE = SL − TL − (NL − DI) − DT. Default target = merchant ship (bulk carrier, AIS 70, 200 m @ 12 kn, JOMOPANS-ECHO decidecade band level) at 6 m; receiver ½ water depth (≤ 30 m), omni (DI 0 — give directivity_index_db for an array). `frequency_hz` is optional: the tool always also evaluates the target's MOST DETECTABLE 1/3-octave band in 50–2000 Hz (`best_band`) — ships are loudest there; a user frequency is reported as asked. If you give `source_level_db` without `frequency_hz`, the range is evaluated at 50 Hz. mode="active" (monostatic, `frequency_hz` required): SE = SL − 2·TL + TS − (NL − DI) − DT, TL from the sonar to the target depth. Defaults: hull sonar at 7 m, SL 220 dB, TS 10 dB (large underwater object, beam aspect), target ½ water depth (≤ 50 m), DI 20 dB (hull array). Noise-limited only: reverberation is NOT included (often the limit in shallow water). NL (unless given) = Wenz ambient noise at the CURRENT forecast wind at the point (

输入参数pdlatlonpfamodemonthplacebearing_degbandwidth_hzfrequency_hzmax_range_kmnoise_level_db另有 9 个

run_bellhop_volume

Bellhop 3D TL volume (stored run)

Full 3D transmission-loss volume around a source (lat/lon or `place`) with Bellhop (all bearings), stored on the platform under a run id. Show `open_url` to the user as a clickable link: it opens this run visualized. Returns the run id, the replication metadata (SSP, seabed, bounding box) and file links (uint8 TL cube .npy + JSON sidecar). Any frequency and radius (keep radius <= ~50 km for reasonable run times).

输入参数latlonmonthplaceradius_kmfrequency_hzsource_depth_m

search_vessels

Search vessels (AIS)

Search live AIS vessels by name or MMSI prefix (global feed).

输入参数limitquery必填

vessels_near

Vessels near a point (AIS)

Live AIS vessels within radius_km of a point (lat/lon or `place`), nearest first, with position, course/speed, type, dimensions and last-update time.

输入参数latlonlimitplaceradius_km

get_vessel

Vessel details + 3D model

Everything about one vessel: live AIS position/track and static particulars, plus its 3D model (unique photo-derived model if generated, else the class archetype) with GLB URL (bow=+Z, up=+Y) and platform links.

输入参数mmsi必填

vessel_source_level

Vessel source level

Underwater radiated-noise source level of a ship (monopole, dB re 1 uPa m, 6 m reference source depth): decidecade spectrum (10 Hz - 100 kHz; model bands to 31.5 kHz, validated 20 Hz - 20 kHz) and broadband level, from the JOMOPANS-ECHO reference spectrum model (MacGillivray & de Jong 2021, J. Mar. Sci. Eng. 9:369), fitted to 1862 measured ship passes. Individual ships scatter about +/-6 dB rms per band. Give an MMSI (live AIS type/speed/length are used) or explicit AIS ship_type / speed / length. vessel_class overrides the AIS mapping (e.g. "vehicle_carrier", which has no AIS type). Missing speed = the class reference speed; missing length = the class mean length of the measured ships. The spectrum is usable directly as `sl_spectrum`.

输入参数mmsibeam_mdraft_mlength_mspeed_knship_typevessel_class

get_vessel_photo

Vessel photo

Photograph of a vessel from Wikimedia Commons, with attribution.

输入参数imommsiname

habitat_received_level

Habitat received level (vessel noise)

Broadband received level (dB re 1 uPa) at a fixed site — a fish farm, reef, hydrophone or marine protected area — from the ships around it, power-summed. Each vessel's source level comes from its AIS class, speed and size; transmission loss is a fast broadband model, so this is a screening estimate (use run_transmission_loss for exact site physics). hours_back = 0 uses the live snapshot; > 0 reconstructs a time series from the AIS archive in 10-minute bins (max/median/quiet levels). Location: lat/lon or `place`. Show `open_url` to the user as a clickable link.

输入参数latlontopplacehours_backmax_range_km

about

About Clairwave

What Clairwave provides, which models/data back each tool, and limits.

无需输入参数。

版本历史

1
  1. v0.4.0最新Sep 16, 2026

评分

已满足适用检查项的 90%

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
  • 已列出工具必需15 分
  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分

    请在仓库或注册中心条目中声明许可证。