AstraEye Labs
officialAI research for U.S. stocks, ETFs and crypto inside Claude or ChatGPT: read your research reports and ask SAGE about them. Every figure is computed from licensed market data and names its source. Remote server, OAuth.
What can you do with AstraEye Labs MCP?
- Read finished reports — Ask your assistant to pull the written summary of one of your finished reports, with its disclosures, via
get_research_report. - Check how a report was built — Have the assistant explain how each analysis in a report works, in the run's own words, using
get_report_methodology. - Browse the analysis catalog — List the analyses AstraEye can run with
list_analyses, or get one in full, including its methodology, withget_analysis. - Find templates and market slices — Explore prebuilt research templates and the market slices a project can cover, such as the S&P 500 or top 100 coins.
- Follow a run — List a project's runs with status and readiness, or check one run's progress, through
list_research_runsandget_research_run. - Ask SAGE to price research — With the SAGE permission on, send a message to your own SAGE chat to price and start research within your daily credit limit.
Hosted MCP Server
npx add-mcp 'https://astraeyelabs.io/mcp'Installs into Claude Code, Codex, Cursor and more
Documentation
AstraEye Labs MCP connector
AI research on U.S. stocks, ETFs and crypto, inside the assistant you already use.
AstraEye Labs is an AI research lab for U.S. stocks, ETFs and crypto. Specialist AI scientists compute every figure in code on licensed market data, and SAGE, the research guide, writes one report from their findings, with the method behind each analysis and the data sources it used.
This repository documents the AstraEye Labs remote MCP server. It lets an AI assistant (ChatGPT, Claude, Gemini, Perplexity, Grok or any MCP client) read a member's finished research reports and their methods, browse the analysis catalog and research templates, follow a run, and ask SAGE to price and start new research.
| Server URL | https://astraeyelabs.io/mcp |
| Transport | Streamable HTTP (stateless, JSON) |
| Sign-in | OAuth 2.1 with PKCE, to the member's own AstraEye Labs account |
| Official MCP Registry | io.astraeyelabs/research |
| Documentation | astraeyelabs.io/research-api |
| For AI assistants | astraeyelabs.io/llms.txt |
What makes the numbers checkable
- Computed, not recalled. Every figure in a report is computed in code from licensed market data and named public sources. The language model explains the results; it does not produce the numbers.
- Method and sources shown. Each report lists its data sources, and each analysis carries its method: what it measures, the rule it applies, and what it cannot support.
- Stocks, ETFs and crypto in one report, from desks for technical structure, fundamentals and macro, news and sentiment, on-chain and alternative data, derivatives, and discovery screens.
- Priced in credits. Ask SAGE what a run costs in Astra Credits before you start it.
Read a published report, no account needed: astraeyelabs.io/r/164j4b202j6w3d204l5z2b53422j54530r0e3528
Guides: Seven checks before you trust an AI's stock or crypto numbers and Stocks, ETFs and crypto in one AI research report.
Connect it
Each connection is made from the assistant: you add the server URL, sign in to AstraEye Labs when it asks, and choose what the assistant may do. A connection is separate from a personal API token and can be disconnected on its own in AstraEye Labs settings.
- Claude: add a custom connector with the server URL
https://astraeyelabs.io/mcp. - ChatGPT: open Plugins, choose the plus button, name it AstraEye, and enter the server URL under Connection. Adding your own plugin can depend on your ChatGPT plan or workspace settings.
- Gemini (web): add AstraEye as a custom app and enter the server URL. Custom apps need a Google AI Pro or Ultra plan.
- Perplexity: open Connectors, choose Custom connector, then Add MCP connector, and enter the server URL. Then open the new connector and choose Add connector. Custom connectors need a Perplexity Pro plan or higher.
- Grok: open Connectors, add a custom connector and enter the server URL.
Who can connect. The connector needs a personal AstraEye Labs account on the Space Analyst plan or above (personal accounts only, 18+, not available in every jurisdiction). There is a free plan, and new accounts start with a 7-day trial with no card. Current plans: astraeyelabs.io/subscribe.
What it returns. The written prose of the member's own reports: the executive summary, the discussion, the conclusion, the prepared questions and answers and, where a report has them, its methodology. It never returns charts, tables or data fields; those stay in AstraEye Labs, and each report links to its page there.
Tools
Read tools (scope research:read):
| Tool | What it does |
|---|---|
list_research_projects | The member's own research projects, newest first |
list_research_reports | The member's finished reports, newest first, with each report's page |
get_research_report | The written summary of one finished report, with its disclosures |
get_report_methodology | How each analysis in a report works, in the run's own words |
list_analyses | The analyses AstraEye can run: scientist, asset class, what it does, the plan it is on |
get_analysis | One analysis in full, including its methodology and what it cannot support |
list_templates | Prebuilt research templates and the analyses each combines |
get_template | One template in full: the analyses it runs, in order |
get_research_project | One of the member's projects, as AstraEye will run it |
list_research_runs | The runs of one project, with status and whether results are ready |
get_research_run | One run's status and progress |
list_market_slices | Market slices a project can cover, such as the S&P 500, the Dow 30, a sector or the top 100 coins |
SAGE tools (scope sage:chat, off by default and chosen on the consent screen):
| Tool | What it does |
|---|---|
send_sage_message | Sends one message to the member's own SAGE chat and returns SAGE's reply. SAGE can price and start research within the daily credit limit the member set for this connection |
get_sage_messages | Reads the member's SAGE chat, including replies that finished later and notices that a run has completed |
Safety and spending
- Sign-in is to the member's own account, with a consent screen and a step-up confirmation. Nothing is shared across accounts.
- By default a connection can only read. With the SAGE chat permission, every reply and every run spends Astra Credits only within the daily credit limit the member chose for that connection.
send_sage_messageis marked as an action, so assistants that ask before actions will ask the member first.- One active connection per assistant. Disconnect it any time in AstraEye Labs settings.
About this repository
This repository holds documentation and the server's registry metadata (server.json) only. The server itself
runs at https://astraeyelabs.io/mcp; its source is not published here.
Support: support@astraeyelabs.io · astraeyelabs.io
AstraEye Labs provides AI-generated research for informational and educational purposes only. Research is impersonal and descriptive, and it is not investment advice. You decide what, if anything, to do. For users 18+; not available in every jurisdiction. AstraEye Labs LLC, Texas.