一个提供美国电力需求和天气数据的MCP服务器。它将数据以MCP工具、资源和提示词的形式暴露出来，通过stdio本地运行，并支持任何MCP客户端：Claude Desktop、Claude Code、LangChain的MCP适配器或MCP Inspector。
数据以一个包含真实观测数据的小型SQLite文件形式提供，因此你可以克隆仓库并直接运行，无需数据库或API密钥。
Python 3.11+ · MCP SDK 1.x (FastMCP) · MIT
MCP是Anthropic提出的一个开放协议。它为AI客户端提供了一种标准方式来调用外部工具和读取外部数据，而不是为每个应用进行自定义集成。服务器声明其提供的内容，客户端连接并使用它。服务器可以提供三类内容：
工具：模型可以调用的函数。
资源：客户端加载作为上下文的只读内容。
提示词：用户可以选择的可重用模板。
此服务器提供了以上全部三类内容。
Related MCP server: euenergy-mcp
工具：
list_regions()：可用的区域，包含行数和日期覆盖范围。
query_demand(region, start, end, limit=168)：某个日期范围内按小时统计的电力需求（单位：MWh）。
daily_demand(region, start, end)：日均、峰值和最低需求。
peak_demand(region, start, end)：单个最高需求的时段。
compare_regions(start, end, regions=None)：跨区域的平均值和峰值。
demand_vs_weather(region, start, end)：日需求数据与该区域温度数据的关联。
资源：
schema://energy：数据库表的纯文本描述。
docs://coverage：当前已加载的区域和日期范围。
提示词：
analyze_demand(region)：一个引导用户通过工具分析某个区域的入门提示。
你需要安装 uv。
git clone https://github.com/visethchapman/energy-mcp
cd energy-mcp
uv run energy-mcp # starts the server on stdio; it waits for a client要在一个界面中浏览这些工具，请使用MCP Inspector：
uv run mcp dev src/energy_mcp/server.py将此添加到 claude_desktop_config.json（在macOS上，路径为 ~/Library/Application Support/Claude/claude_desktop_config.json），然后重启Claude Desktop：
{
"mcpServers": {
"energy": {
"command": "uv",
"args": ["--directory", "/absolute/path/to/energy-mcp", "run", "energy-mcp"]
}
}
}claude mcp add energy -- uv --directory /absolute/path/to/energy-mcp run energy-mcp然后使用自然语言提问，客户端会自动选择相应的工具：
“2024年7月，美国哪个电网的平均需求最高？”
“2024年ERCOT的峰值出现在什么时候？当天休斯顿有多热？”
“比较CISO和PJM在1月第一周的需求。”
客户端调用工具时，从捆绑数据中返回的真实结果：
> peak_demand(region="ERCO", start="2024-01-01", end="2024-12-31")
{ "region": "ERCO", "period": "2024-08-20 23:00:00", "demand_mwh": 85544.0 }
> compare_regions(start="2024-07-01", end="2024-07-31")
[ {"region":"PJM", "avg_mwh":109674.4, "peak_mwh":153121.0},
{"region":"ERCO", "avg_mwh": 60200.6, "peak_mwh": 81200.0},
{"region":"CISO", "avg_mwh": 32008.4, "peak_mwh": 44727.0},
{"region":"NYIS", "avg_mwh": 21494.1, "peak_mwh": 28990.0} ]
> demand_vs_weather(region="ERCO", start="2024-07-01", end="2024-07-03")
[ {"day":"2024-07-01", "avg_mwh":67736.3, "tmax_c":37.8, "tmin_c":26.7},
{"day":"2024-07-02", "avg_mwh":68463.9, "tmax_c":36.1, "tmin_c":27.2},
{"day":"2024-07-03", "avg_mwh":68017.8, "tmax_c":34.4, "tmin_c":26.7} ]demand_vs_weather 是一个有趣的工具。ERCOT的需求随着夏季休斯顿温度的升高而上升（空调用电负荷），该工具会同时返回这两个序列，以便模型能够指出它们之间的关联。
所有数据均为真实数据，并已截取至2024日历年，以确保文件大小保持在约5 MB。
表名 | 来源 | 行数 |
| EIA开放数据API，4个平衡机构：ERCO、CISO、PJM、NYIS | ~35,000 |
| NOAA GHCN-Daily，每个区域主要负荷区附近的一个气象站 | ~1,500 |
| NOAA气象站元数据 | 4 |
| 数据表文档 | 3 |
SQLite文件已包含在仓库中，因此运行时无需获取任何数据。若要重新构建它或扩大日期范围（需要免费的EIA API密钥）：
uv run --extra build python scripts/build_sample_db.py --start 2023-01-01 --end 2024-12-31设置 ENERGY_MCP_DB 环境变量，以使服务器指向不同的SQLite文件。
只读。数据库以SQLite的 mode=ro 模式打开，因此没有工具能够写入。
参数化SQL。参数作为绑定变量传递，而非直接格式化到查询字符串中，因此无法注入SQL。
输入验证。区域和日期参数在到达数据库之前会进行校验，错误的输入会返回清晰的错误信息，供模型处理。
无需密钥即可运行。数据包含在仓库中；EIA密钥仅用于重新构建数据。
uv run pytest
uv run mcp dev src/energy_mcp/server.py项目结构：
src/energy_mcp/
server.py # FastMCP server: tools, resources, prompts
db.py # read-only SQLite access
data/energy_sample.sqlite # bundled real data
scripts/build_sample_db.py # how the sample was assembled
tests/test_tools.py该项目使用MCP 1.x SDK（FastMCP API），这也是当前Claude Desktop文档和 mcp CLI所使用的版本。2.x SDK是一个较新的重新设计版本。
energy-text2sql 是一个基于相同EIA和NOAA数据的文本转SQL智能体。相同的数据，不同的接口：前者通过编写SQL来回答问题，而本项目则将数据作为MCP工具提供。
MIT
Compare average and peak demand across regions over the same date range. Useful for questions like "which grid was busiest last summer?".
Args:
start: Inclusive start date, 'YYYY-MM-DD'.
end: Inclusive end date, 'YYYY-MM-DD'.
regions: Optional subset of region codes. Defaults to all regions.
| Name | Required | Description | Default |
|---|---|---|---|
| end | Yes | ||
| start | Yes | ||
| regions | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description does not disclose behavioral traits such as data aggregation details, performance implications for many regions, or any side effects. Since no annotations are provided, the description carries the full burden but only provides basic usage info.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise, with a clear purpose and example, followed by parameter details. The example is helpful, though the 'Args' section could be integrated more seamlessly.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given there is an output schema and sibling tools exist, the description adequately distinguishes the tool and explains inputs. It could be more complete by mentioning that the output includes both average and peak metrics, or by noting that region codes can be obtained via list_regions.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description explains the start and end parameters with date format and inclusiveness, and clarifies the optional regions parameter defaults to all. However, schema coverage is 0%, so the description compensates partially but does not specify accepted region codes or error handling for invalid inputs.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states it compares average and peak demand across regions over a date range, with a specific example of use, clearly distinguishing it from siblings like peak_demand or daily_demand.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides an example question to indicate when to use the tool, but does not explicitly state when not to use it or mention alternative tools for single-region queries.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Return daily demand statistics (average, peak, minimum MWh) for one region over a date range. Use this instead of query_demand for ranges longer than a few days.
Args:
region: Balancing authority code.
start: Inclusive start date, 'YYYY-MM-DD'.
end: Inclusive end date, 'YYYY-MM-DD'.
| Name | Required | Description | Default |
|---|---|---|---|
| end | Yes | ||
| start | Yes | ||
| region | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so description carries full burden. It states the output includes average, peak, and minimum MWh, likely aggregated daily. However, it does not disclose whether the tool is read-only, whether it handles time zones, or what happens if data is missing for a date range. These gaps limit transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences plus a parameter list, all relevant. It front-loads the purpose and usage recommendation. The parameter list is typical for Google-style docstrings but adds a bit of vertical space. No wasted sentences.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given 3 simple string parameters, an output schema exists (presumably documenting return fields), and no nested objects, the description is largely complete. It provides purpose, usage, and parameter formats. Missing details like valid region codes or date range limits are minor given the output schema's presumed coverage.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 0%, so the description must compensate. It explains region as a balancing authority code, start and end as inclusive dates with 'YYYY-MM-DD' format. This adds significant meaning beyond the bare schema property titles. However, it does not specify valid region codes or constraints on date range length.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool returns daily demand statistics (average, peak, minimum MWh) for a region over a date range. It distinguishes itself from the sibling query_demand by explicitly recommending use for ranges longer than a few days.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Provides explicit guidance on when to use this tool over query_demand ('for ranges longer than a few days') and lists three parameters with format hints. No exclusions or alternatives are elaborated beyond query_demand, but the recommendation is specific and actionable.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Join daily average demand against daily temperature for a region, using the NOAA weather station nearest that region's main load zone. Reveals the demand-temperature relationship (heating and cooling load).
Args:
region: Balancing authority code.
start: Inclusive start date, 'YYYY-MM-DD'.
end: Inclusive end date, 'YYYY-MM-DD'.
Returns one row per day: average demand (MWh) and that day's max/min temp (C).
| Name | Required | Description | Default |
|---|---|---|---|
| end | Yes | ||
| start | Yes | ||
| region | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description bears full responsibility for transparency. It discloses the method ('using the NOAA weather station nearest that region's main load zone'), the output structure (one row per day, average demand MWh, max/min temp C), and the join logic. It does not mention potential limitations like missing data or time zone handling, but the core behavior is well explained.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise and well-structured: a lead sentence explaining the purpose, followed by a bullet-like Args list. Every sentence adds value, and there is no redundancy. The key information is front-loaded.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's moderate complexity (3 required parameters, output schema exists), the description is complete. It explains the input format, the join logic, and the output structure (one row per day, demand and temperature). The presence of an output schema means the full return format is already defined, so the description's summary is sufficient.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description provides clear, actionable semantics for all three parameters. It explains region as 'Balancing authority code', start and end as 'Inclusive start date, YYYY-MM-DD' and 'Inclusive end date, YYYY-MM-DD'. This adds significant value beyond the schema, which only has titles. With 0% schema description coverage, the description fully compensates.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: joining daily average demand against daily temperature for a region to reveal the demand-temperature relationship. It uses a specific verb ('join') and resource ('demand vs weather') and distinguishes itself from sibling tools like query_demand, daily_demand, and compare_regions by focusing on the combined analysis.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explains the tool's use case (revealing heating and cooling load relationships) but does not explicitly state when to avoid it or name alternatives. However, the context (sibling tools) and the description's wording imply that this tool is for combined demand-weather analysis, which is a distinct purpose among the siblings.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List the balancing authorities available, with row counts and the date range of demand data for each. Call this first to see what you can query.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, so description carries the burden. It clearly indicates a read operation (listing), no side effects mentioned but context implies safe discovery. Could be more explicit about read-only nature, but adequate for a listing tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences: first explains output, second provides usage guidance. Front-loaded, no wasted words, each sentence earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a zero-parameter listing tool with an output schema, the description tells what is returned (balancing authorities, row counts, date ranges) and when to use it. Complete for the agent to understand and invoke correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
No parameters exist, and schema coverage is 100% trivially. Description does not need to add parameter info. Baseline 4 is appropriate as no value beyond schema is necessary.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Description clearly states the tool lists balancing authorities with row counts and date ranges. Distinguishes from siblings by specifying the output content (row counts, date ranges) and by positioning it as a discovery tool to be called first.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly says 'Call this first to see what you can query,' providing a clear when-to-use instruction and implying it is a prerequisite for other tools like query_demand or daily_demand.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Find the single highest-demand hour for one region over a date range, returning the timestamp and the demand in MWh.
Args:
region: Balancing authority code.
start: Inclusive start date, 'YYYY-MM-DD'.
end: Inclusive end date, 'YYYY-MM-DD'.
| Name | Required | Description | Default |
|---|---|---|---|
| end | Yes | ||
| start | Yes | ||
| region | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description fully carries the burden of behavioral disclosure. It states that the tool returns a single timestamp and demand value, which is clear. However, it does not cover edge cases (e.g., no data in range, invalid region), error handling, or any mutation aspects. For a simple read-only query, this is adequate but not thorough.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is extremely concise: one purpose sentence followed by a parameter list. Every sentence is necessary and there is no fluff. The key information is front-loaded in the first sentence.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the simplicity of the tool (3 string parameters, no output schema, no annotations), the description covers purpose, parameters, and return value. It is complete enough for an agent to use correctly. Missing details like error handling or behaviour when no data exists are minor gaps for such a straightforward query.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 0%, so the description must explain parameters. It does: region is a 'balancing authority code', start and end are 'inclusive start/end date' with format 'YYYY-MM-DD'. This adds meaningful context beyond the schema's plain string type. However, it could be improved by providing examples or a list of valid region codes.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb 'Find' and identifies the resource as 'single highest-demand hour', with clear scope of 'one region over a date range'. It also mentions the return values (timestamp and demand in MWh). This clearly distinguishes it from sibling tools like query_demand (general data) and daily_demand (per-day peaks).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to use this tool versus alternatives. There is no mention of when it is appropriate, when not to use it, or how it compares to sibling tools like query_demand, daily_demand, or compare_regions. The context of 'one region' is stated but not contrasted with multi-region tools.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Return hourly electricity demand (MWh) for one region over a date range.
Args:
region: Balancing authority code (e.g. ERCO, CISO, PJM, NYIS).
start: Inclusive start date, 'YYYY-MM-DD'.
end: Inclusive end date, 'YYYY-MM-DD'.
limit: Max rows to return (default 168 = one week of hours).
Returns hourly rows ordered by time. For long ranges prefer daily_demand.
| Name | Required | Description | Default |
|---|---|---|---|
| end | Yes | ||
| limit | No | ||
| start | Yes | ||
| region | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Since no annotations exist, the description carries full burden. It discloses sorting behavior ('Returns hourly rows ordered by time'), the default limit meaning (168 = one week of hours), and the effect of using long ranges (prefer daily_demand). It does not mention pagination behavior, whether the limit is a hard cap or default, or what happens if range exceeds limit, but the essential behavioral traits are adequately covered given the tool's simplicity.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise at under 70 words, front-loaded with the core purpose and unit (MWh), then uses a structured Args list for parameters, followed by behavior and usage guidance. Every sentence serves a distinct purpose, and there is no redundancy or fluff.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has 4 parameters (3 required, 0 enums, 0% schema coverage) and an output schema, the description appropriately covers input semantics, behavior (ordered hourly rows), and usage guidance with sibling reference. With no annotations, it compensates well. The output schema presumably describes return fields, so the description's lack of explicit return format is acceptable.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must compensate. It provides clear, helpful semantics for each parameter: region meaning (Balancing authority code with examples), start/end format ('YYYY-MM-DD') with 'inclusive' clarified, and limit meaning (max rows, default 168 = one week of hours). This adds substantial value beyond the schema labels (e.g., 'Region', 'Start', 'End', 'Limit').
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description explicitly states the tool returns hourly electricity demand (MWh) for one region over a date range, using a specific verb 'Return' and identifying the resource (demand data), region, and output unit (MWh). It also distinguishes itself from siblings like daily_demand and peak_demand by mentioning hourly data and the limit default of 168 rows (one week of hours), and directly points to daily_demand as an alternative for long ranges.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives clear when-to-use guidance: 'For long ranges prefer daily_demand', explicitly naming the sibling alternative. It also defines the purpose (hourly data for one region) implicitly telling the agent not to use this for multiple regions or aggregated summaries (compare_regions, peak_demand). The Args docstring reinforces required parameters and their formats.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0compare_regionsdaily_demanddemand_vs_weatherlist_regionspeak_demandquery_demandScored across 6 tools
Each tool serves a distinct purpose: listing regions, retrieving hourly/daily/peak demand, comparing across regions, and correlating demand with weather. No functional overlap.
All six tool names follow a consistent verb_noun pattern with snake_case: list_regions, query_demand, daily_demand, peak_demand, compare_regions, demand_vs_weather. Naming is predictable and clear.
With 6 tools, the server is well-scoped for its purpose – covering region discovery, multiple demand granularities, inter-region comparison, and weather integration. No bloat or deficiency.
Core demand querying workflows are fully supported. A minor gap is the lack of a raw hourly multi-region query (compare_regions only returns averages/peaks). Overall, the surface is thorough for typical use cases.
MCP server for building and testing AI agents with multi-model experimentation and insights.
Hosted MCP server for live public-data APIs and Skills for AI agents.
MCP server for querying and analyzing data from ad platforms, analytics tools, and spreadsheets
Hosted MCP server exposing US hospital procedure cost data to AI assistants
