Who it is for
Energy analysts who need US generation figures with a traceable source for every number.
AI Connectors · Data & Analytics
Source-cited US power generation and capacity data over MCP
Overview
exascale.build publishes US power system data drawn from EIA filings: operating, planned and retired generation capacity from EIA-860M, net generation from EIA-923, and capacity factor. Takezo's stated design point is provenance. Every value returns with its source and an as-of date, and any figure resolves to the raw source cell with a matching SHA-256 hash.
Each dataset is reachable two ways, a describe call that states the shape and scope of a capability and a query call that runs against it. The listing says tools declare what they will not answer, so an agent refuses out-of-scope questions instead of guessing. Access is over MCP or REST without credentials, and the connector is community-listed.
Energy analysts who need US generation figures with a traceable source for every number.
Fourteen tools answered against eight claimed in the listing. They pair up, a describe_ call and a query_ call for each dataset across power capacity, generation, asset ownership, fuel cost and plant costs, with list_capabilities_v1 and get_source_evidence_v1 covering provenance.
Availability
exascale.build is listed in the Claude connectors directory only. Takezo has no app under the same domain in the ChatGPT directory; 392 of the 3,150 products here are on both marketplaces, and this is not one of them.
The MCP server behind a Claude connector is usually the same server a ChatGPT app would need, so the absence is a distribution decision rather than a technical one.
Claude listing
exascale.build publishes 8 tools to Claude. That is a mid-sized surface: 37.7% of the 1,253 Claude-listed connectors expose fewer.
The median Claude-listed connector filed under Data & Analytics publishes 12, so this one sits 4 tools below its bucket's median, 123rd of 198.
It falls in the 7-10 tools band, which holds 14.6% of the Claude directory.
A direct call to the endpoint returned 66 tools, 58 more than the 8 the listing publishes. Probed 2026-08-07.
list_capabilities_v1get_source_evidence_v1describe_capability_v1query_capability_v1describe_power_capacity_v1query_power_capacity_v1describe_power_generation_v1query_power_generation_v1describe_power_asset_ownership_v1query_power_asset_ownership_v1describe_power_fuel_cost_v1query_power_fuel_cost_v1describe_power_plant_costs_v1query_power_plant_costs_v1describe_natural_gas_prices_v1query_natural_gas_prices_v1describe_power_demand_v1query_power_demand_v1describe_power_demand_rollup_v1query_power_demand_rollup_v1describe_power_retail_sales_v1query_power_retail_sales_v1describe_power_interconnection_queue_v1query_power_interconnection_queue_v1describe_power_interconnection_queue_pjm_v1query_power_interconnection_queue_pjm_v1describe_power_interconnection_queue_pjm_cycle_v1query_power_interconnection_queue_pjm_cycle_v1describe_power_interconnection_queue_caiso_v1query_power_interconnection_queue_caiso_v1describe_power_interconnection_queue_nyiso_v1query_power_interconnection_queue_nyiso_v1describe_power_interconnection_queue_isone_v1query_power_interconnection_queue_isone_v1describe_power_interconnection_queue_ercot_v1query_power_interconnection_queue_ercot_v1describe_power_interconnection_queue_spp_v1query_power_interconnection_queue_spp_v1describe_power_price_ercot_v1query_power_price_ercot_v1describe_power_capacity_accreditation_pjm_v1query_power_capacity_accreditation_pjm_v1describe_power_capacity_market_pjm_v1query_power_capacity_market_pjm_v1describe_power_capacity_accreditation_miso_v1query_power_capacity_accreditation_miso_v1describe_power_capacity_market_miso_v1query_power_capacity_market_miso_v1describe_ai_infrastructure_construction_v1query_ai_infrastructure_construction_v1describe_ai_infrastructure_employment_v1query_ai_infrastructure_employment_v1describe_ai_infrastructure_trade_v1query_ai_infrastructure_trade_v1describe_ai_infrastructure_equipment_trade_v1query_ai_infrastructure_equipment_trade_v1describe_ai_infrastructure_production_v1query_ai_infrastructure_production_v1describe_robotics_trade_v1query_robotics_trade_v1describe_robotics_adoption_v1query_robotics_adoption_v1describe_space_satellite_filings_v1query_space_satellite_filings_v1describe_capacity_factor_v1query_capacity_factor_v1exascale.build publishes a remote MCP endpoint. Claude connects to it over HTTP; there is nothing to install locally.
Every connector in the Claude directory is remote, all 1,253 of them, so transport is not a differentiator there.
https://api.exascale.build/mcp exascale.build is community-listed in the Claude directory, the tier 770 of 1,253 connectors sit in (61.5%). It is the default for a self-published server, and says nothing either way about quality.
In the publisher's words
Agent-first, source-cited data layer for the US power system. Query operating, planned, and retired generation capacity (EIA-860M), net generation (EIA-923), and capacity factor — over MCP or REST. Every value returns with its source and as-of date, and any number resolves to its raw source cell with a matching SHA-256, so an agent's answer is verifiable in one call. Tools declare what they won't answer, so the agent refuses out-of-scope questions instead of guessing. Free and open today; industrial and logistics rolling out.
Quoted from the connector's own Claude directory listing as captured in August 2026. Node8 did not write it and does not vouch for it.
Nearby
Web analytics questions answered with session-level evidence
Claude
Materials characterisation data, notebooks, and instrument streams
Claude
Read-only energy analytics for commercial and industrial sites
Claude
Enterprise data context: search, lineage, ownership and governance
Claude
Browse, query and run dbt projects and the semantic layer
Claude
Notebook workspace access for analysis and execution
ChatGPT
This is an independent catalogue entry compiled by Node8 from exascale.build's public marketplace listing as it stood in August 2026. Listing facts come from the marketplaces; the comparisons against the rest of the directory and any live-endpoint measurements are Node8's. Node8 is not affiliated with Takezo, and inclusion here is not an endorsement.
Node8 builds MCP servers end to end: deciding which tools are worth exposing, the authentication and permission model, the review submissions, and the onboarding that gets them used. One server serves ChatGPT, Claude, and Microsoft Copilot.