Nameplate Data

Search Projects

Search the data-center project portfolio by name, alias, or county. This is the search-first entry point: start here to resolve a project_id before calling get_project_dossier or compare_projects. Searches the 364-project portfolio across 30 states. Args: query: substring matched against project name, county, and state (e.g. "Stargate", "Armstrong", "Google"). Optional -- omit to list the whole (filtered) portfolio.

Get Project Dossier

Return the full evidence bundle for one project. The complete underwriting file: summary, power/interconnect evidence, cost estimates (with source URLs), financing, regulatory filings, timeline events, evidence depth-tier criteria, latest risk scores with explanation bullets, and this project's holdout findings (evidence we refused to count). Every evidence row cites its source_id. Get the project_id from search_projects first. Args:

Query Holdouts

Query the refusal trail -- evidence we explicitly refused to count. This is the differentiator. Where other data feeds only report what they booked, this surfaces what we reviewed and deliberately did NOT promote to coverage, each with a recorded reason and a do-not-promote decision: a capex number a press release implied but never disclosed cleanly, an MW figure that mixed PPA capacity with load, an entity link we could not substantiate. Use it t

Query Cost Estimates

Query booked cost evidence across the portfolio. Returns the latest cost estimate per project with its basis, confidence, benchmark label, and source URL -- the receipted capex signal. Args: state: two-letter state filter (e.g. "TX"). estimate_type: filter on the estimate type string (e.g. "developer_market_tracker_estimate"). min_amount: minimum estimate amount in USD. limit: max rows to return (default 25

Compare Projects

Compare 2-5 projects column-aligned for an investment-committee view. Puts MW, cost, computed $/MW, depth tier, risk dimensions, peer group, and evidence readiness side by side. Resolve the project_ids with search_projects first. Args: project_ids: 2 to 5 project identifiers to compare. Returns: JSON with a ``projects`` list, one column per project.

State Rollup

Roll up a state's portfolio with the ISO/utility queue backdrop. Returns project counts by status and evidence depth tier, announced / queue-matched / energized MW, cost coverage, and the booked ISO/utility aggregate queue context (e.g. the ~410 GW ERCOT large-load figure for TX) with a citable source. Puts a state's numbers against the grid-scale backdrop. Args: state: two-letter state code (e.g. "TX"). Returns:

Market Rollup

Roll up the whole tracked portfolio into a market view. Returns totals (project count, disclosed-MW capacity, MW-weighted average investability) plus breakdowns by status, state, risk band, probability-score histogram, rolled-up financing category, and expected completion year. Every breakdown carries the disclosed-vs-undisclosed MW split explicitly -- undisclosed MW is reported as absent, never imputed. Returns: JSON rollup with tota

Search Sources

Full-text search the approved source corpus for grounding context. Searches the news reports, filings, and government records that back the product's claims, and returns highlighted excerpts -- never full documents -- with each source's URL and related project ids, so an answer can cite receipts. Use ``source_id`` to pull the context behind a citation another tool returned (cost estimates, holdouts, and dossiers all carry source ids). Only approve

Get Access Info

Report the caller's tier, rate-limit state, and what each tier unlocks. Introspection with no side effects and no DB access. This tool answers for any caller (including anonymous / free / pro) so a misconfigured client can diagnose its access; every other tool is provided under Team agreements over the HTTP transport. Returns: JSON with the caller's tier, remaining rate-limit budget, and the full tier table (free = public workbenc