Atlasresearch

Workspace

WorkspacePokedexHPC MapPortfolioCalendarTaiwan RevenueMemory
Calculators
  • 01AI Factory
  • 02CPO Stack
  • 03HBM Bridge
  • 04GPU Cloud TCO
  • 05Inference Tokens
  • 06Power Bottleneck
  • 07Custom ASIC
  • 08OCS TAM
  • 09N3 Allocation
  • 10Optical TAM
Agent APISubscribe
Sign in
Agent API Docs
Atlasresearch

Workspace

WorkspacePokedexHPC MapPortfolioCalendarTaiwan RevenueMemory
Calculators
  • 01AI Factory
  • 02CPO Stack
  • 03HBM Bridge
  • 04GPU Cloud TCO
  • 05Inference Tokens
  • 06Power Bottleneck
  • 07Custom ASIC
  • 08OCS TAM
  • 09N3 Allocation
  • 10Optical TAM
Agent APISubscribe
Sign in
Agent API/Docs

Agent API docs

API Docs

Use this page to wire agents to Atlas endpoints.

Base URL

/api/agent/v1

Subscribe or sign in with your Substack email to create keys.

Subscribe to Jason's ChipsSign in

Run calculator scenarios

Fetch calculator definitions, change assumptions, run the shared Atlas math, and compare revenue, margin, payback, power, or capacity sensitivity.

Build company read-throughs

Combine company profiles, taxonomy membership, calendar events, Taiwan revenue, memory data, and dashboard rows before drafting a financial note.

Monitor catalyst windows

Watch earnings, transcript availability, revenue releases, and source updates, then refresh analysis when new data changes the debate.

Preserve source discipline

Use structured endpoints and generated timestamps so agents can separate Atlas-authored context, live data, and calculator output.

Agent financial analysis loop

Agent Workflow

  1. 1Discover endpoints through the catalog so the agent knows which datasets and calculator models are available.
  2. 2Pull company, taxonomy, calendar, memory, or Taiwan revenue context for the relevant AI infrastructure theme.
  3. 3Run one or more calculator scenarios with explicit assumptions instead of free-form spreadsheet guessing.
  4. 4Compare output deltas, flag the assumption that matters most, and draft a concise investment read-through.

Example

Run Atlas calculator math from an agent

curl -X POST /api/agent/v1/calculators/ai-factory/run \ -H "Authorization: Bearer atlas_..." \ -H "Content-Type: application/json" \ -d '{"inputs":{"activeMw":120,"gpuTdpWatts":1200,"rackKw":120}}'

Agents can vary inputs such as active MW, rack density, accelerator power, ASP, HBM stacks, wafer starts, or utilization, then summarize which assumption drives the largest change.

Endpoint reference

Call the catalog endpoint first when wiring an agent. It returns this same capability map with method, path, cache behavior, and access status.

11 capabilities
MethodPathStatusCache
GET
/api/agent/v1/catalog

Discover available Atlas agent API capabilities.

availableshort
GET
/api/agent/v1/taxonomy

AI infrastructure categories, layers, segments, and labels.

availablelong
GET
/api/agent/v1/companies?category=&segment=&marketCap=&q=&limit=

Atlas AI infrastructure company universe with optional category, segment, market cap, and text filters.

availablemedium
GET
/api/agent/v1/companies/{slug}

Atlas-authored company profile, taxonomy memberships, and peer set.

availablemedium
GET
/api/agent/v1/companies/{slug}/dashboard

Company dashboard data: quote, financials, earnings, transcript dates, ownership, filings, and news.

guardedshort
GET
/api/agent/v1/companies/{slug}/transcripts/{fyYYYY-qN}

Authenticated access to a single earnings-call transcript for agent analysis, subject to upstream data terms.

guardedmedium
GET
/api/agent/v1/calendar

Earnings, transcript, and Atlas catalyst calendar for the AI infrastructure universe.

availableshort
GET
/api/agent/v1/taiwan-revenue

Official Taiwan monthly revenue tracker, baskets, leaders, laggards, and source statuses.

availablelong
GET
/api/agent/v1/memory?limit=

Persisted memory spot-price observations and memory-company financial rows.

availableshort
GET
/api/agent/v1/calculators

Available Atlas calculator models, presets, and default inputs.

availablelong
POST
/api/agent/v1/calculators/{calculator}/run

Run shared calculator math for AI factory, CPO stack, HBM revenue bridge, GPU cloud TCO, custom ASIC economics, OCS TAM, N3 allocation bottlenecks, optical transceiver TAM, inference token economics, or data center power bottleneck scenarios.

availableno-store