StockLens

Vị thế của một công ty trên thị trường và lý do. Các công ty lớn nhất trong mỗi thị trường không cần tài khoản.

Máy chủ MCP được lưu trữ

npx add-mcp 'https://mcp.stocklens.com/mcp'

Cài vào Claude Code, Codex, Cursor và nhiều công cụ khác

Tài liệu

AI Integrations

Where a company stands in its market, and why, with the evidence for and against, from the ChatGPT Plugin, the Claude Connector or any MCP-compatible client. Protected tools sign in with OAuth.

Start free

Where StockLens runs

One protocol, the assistants you already use

StockLens exposes the same typed MCP tools to every compatible client, whether you connect through the ChatGPT Plugin, the Claude Connector, or your own MCP client.

ChatGPT Plugin

Ask StockLens for scores, analysis and screens inside ChatGPT.

View StockLens in the ChatGPT plugin directory

Claude Connector

Add StockLens to claude.ai and use the same tools in the conversation.

View StockLens in Claude's connector directory

Also use it as a direct MCP server

Point any MCP-compatible client at the StockLens server. Point it at https://mcp.stocklens.com/mcp, published in the Official MCP Registry as com.stocklens/stocklens.

Glama

Capability mirror

The tools your agent actually calls

The complete StockLens MCP surface has three public tools and eight OAuth-protected tools.

ping

Auth: none

Verify that the StockLens MCP connection is healthy without reading account or market data.

Outputstatus: ok, text: pong


get_market_summary

Auth: none

Retrieve the latest nightly analysis summary for a scored company: the score and rating, the domain breakdown, the drivers that shaped the reading, its standing within its own market where that market has one, and the date it was computed.

Inputs

tickerstring (e.g. AAPL, MSFT, TSLA)

exchange?string (e.g. NASDAQ, TSX, LSE). Auto-resolved via autocomplete if omitted.

Outputstatus; when scored: composite_score, rating (nullable), domain_scores[], key_drivers[], lens, domain_coverage, standing, analysis_date, freshness, config_context, reading_ref. Any other status is an answer carrying message and next_step


get_shortlist_summary

Auth: none

Read the companies the person names in one call. Each scored company carries its complete nightly reading and, where its market has one, its standing: one of five bands within that company's own market, named with the market and its date. Companies from different markets are read side by side, never on one scale. It does not sort the market, start an analysis or change a portfolio.

Inputs

companiesarray of 1 to 10 { ticker, exchange } -- the person's own list, in their order

Outputcompanies[] (one get_market_summary result per company, in the order given, each with its own standing and date), standing_definition (once per call), scope: named_shortlist


inspect_reading

Auth: OAuth

Explain an exact saved reading from the evidence it already holds: why a score reason, finding, qualification or assessment says what it does. It starts no analysis, uses no credit and writes nothing.

Inputs

reading_refstring (the exact reading_ref a StockLens result returned)

view?overview | finding | assessment | observation (default overview)

selector?string (copied from the result's inspection arguments; needed for every view except overview)

Outputstatus, scope, reading_ref, view, kind, as_of, company, availability, verdict, assessment, finding, selected_reason, evidence[], inspection[], narrative, limitations[], message, report_url


initiate_analysis

Auth: OAuth

Kick off a full StockLens AI analysis for a supported ticker. A new run may asynchronously consume one analysis credit.

Inputs

tickerstring (e.g. AAPL)

exchange?string (e.g. NASDAQ, TSX, LSE). Auto-resolved via autocomplete if omitted.

time_range?Intended holding period from the analysis date: short_2_wk | medium_1_mo | long_1_yr (default medium_1_mo). Shorter horizons give recent price action, momentum and news more emphasis than longer ones; longer horizons give business fundamentals and quality more emphasis. Not a prediction of price or returns.

analysis_focus?low | medium | high (default medium).

depth?standard | plus (default standard)

Outputanalysis_id, status, estimated_seconds, depth, credit_type, new_run_started, reused_existing, result (when the scores are ready)

CreditsLENS_AI``LENS_AI_PLUS


fetch_analysis_result

Auth: OAuth

Check an analysis by id. Each call returns the current state: progress while it runs, then the saved report as soon as it exists. The written summaries can arrive after the saved report, and the result says whether they are still being written, are ready, or could not be completed.

Inputs

analysis_idstring (id returned by initiate_analysis)

Outputstatus, progress_pct, retry_after_seconds -- or the saved report: result (symbol, analysis_focus, analysis_focus_display, display_composite_score, display_rating, evidence, domains, key_drivers, lens, domain_coverage), base_digest, report_url, narrative (status, revision) and message; synthesis_report, key_strengths, key_challenges and each domain's analysis are included once narrative.status is ready


get_portfolio_insights

Auth: OAuth

Read one of the authenticated user's StockLens portfolios, the default one unless a portfolio is named: every company, holding or watching, with its score, rating word, market and Fundamental standing within that market, its largest sector share and its biggest movers over 7 nights, each naming its market. A portfolio whose companies are all listed in one market also gets that market's average score and rating word and its change over 7 and 30 nights; across markets there is no average, because scores from different markets are not on one scale. The whole portfolio is returned, up to the plan's capacity.

Inputs

portfolio_id?string (from the portfolios list)

portfolio_name?string (exact name, case does not matter)

Outputstatus (ok, empty or portfolio_not_found), portfolio_id, portfolio_name, is_default, portfolio_count, portfolios[] (when there is more than one), caption, trust[], company_count, average_composite_score, rating, window_change (1w, 1m) and average_market (these four only for a one-market portfolio), scale_note (across markets), markets[], concentration, top_movers[], companies[] (each with market and standing), tier_limited, as_of


add_to_watchlist

Auth: OAuth

Add one supported company to one of the authenticated user's StockLens portfolios, the default one unless a portfolio is named; the result names the portfolio that received it. Following only, one ticker per call: it records no ownership, writes no holding quantity or transaction, and places no trade. Free plans follow 10 stocks across Watchlist and Holdings; Pro and Max plans are unlimited. At the Free capacity nothing is added and the result states the limit.

Inputs

tickerstring (e.g. AAPL) -- one ticker, not a list

exchange?string (e.g. NASDAQ, TSX, LSE). Auto-resolved via autocomplete if omitted.

portfolio_id?string (from the portfolios list)

portfolio_name?string (exact name, case does not matter)

Outputstatus (added, already_in_portfolio, over_capacity, ambiguous, not_found or portfolio_not_found), message, ticker, exchange, company_name, portfolio_id, portfolio_name, is_default, portfolio_count, company_count, plan_limit (absent when the plan is unlimited), scope: user_watchlist_update


remove_from_watchlist

Auth: OAuth

Remove one company from one of the authenticated user's StockLens portfolios, the default one unless a portfolio is named; the result names the portfolio. It removes that entry and nothing else; an entry that already carries transactions is refused rather than deleted, so transaction history is never discarded by an agent.

Inputs

tickerstring (e.g. AAPL) -- one ticker, not a list

exchange?string (e.g. NASDAQ, TSX, LSE). Auto-resolved via autocomplete if omitted.

portfolio_id?string (from the portfolios list)

portfolio_name?string (exact name, case does not matter)

Outputstatus (removed, not_in_portfolio, has_transactions, ambiguous_listing or portfolio_not_found), message, ticker, exchange, portfolio_id, portfolio_name, is_default, portfolio_count, company_count, scope: user_watchlist_update


add_to_watchlist_batch

Auth: OAuth

Add a list of companies the user gives, up to 100, to one of the authenticated user's StockLens portfolios, the default one unless a portfolio is named, in two calls: a preview that adds nothing and states every row's outcome, then a commit of only the rows the user approves, bound to that preview. Following only: it records no ownership, writes no holding quantity or transaction, and places no trade. Each row is held to the same portfolio capacity as add_to_watchlist.

Inputs

rowsarray of 1 to 100 { query, exchange } -- one company per row, a ticker or a name

confirmfalse previews and adds nothing | true adds the approved rows

preview_id?string (from the preview; needed when confirm is true)

confirmed_indexes?array of the row indexes the user approved (needed when confirm is true)

portfolio_id?string (from the portfolios list)

portfolio_name?string (exact name, case does not matter)

Outputstatus (previewed, committed, portfolio_not_found or preview_mismatch), preview_id, portfolio_id, portfolio_name, is_default, portfolio_count, company_count, summary, summary_line, capacity, message, rows[] (index, input, outcome, label, listing, candidates, message), scope: user_watchlist_batch


discover_stocks

Auth: OAuth

Screen the companies StockLens scores nightly in one listing market by exchange, sector, industry, market-cap band, and volume, then sort by 6 StockLens scores, market cap, volume or ticker. Every list is computed inside one market, the one the request names or otherwise the United States, and the result names it; lists from different markets are never merged. Popularity ordering is unavailable.

Inputs

sort_by?composite_score | fundamental_score | technical_score | risk_score | sentiment_score | macro_score | ticker | market_cap | volume (default composite_score)

direction?top | bottom (default top)

market_cap_band?MEGA | LARGE | MID | SMALL | MICRO

volume_floor?100K | 500K | 1M | 5M

macro_in / exchange_in / sector_in / industry_in?arrays restricting the screen; macro_in or exchange_in names its one market, and a request spanning markets returns multiple_markets with nothing ranked

limit?1-100 (default 10)

Outputstatus, market, market_name, rows[] (scores present only when unlocked), scores_locked, access_tier, message, sort_substituted, reach_capped_at, rows_returned, effective_sort, effective_direction, as_of_date, data_source: nightly_scores

Coverage

What StockLens scores

StockLens scores selected listings nightly across North America, Europe and Asia-Pacific; macro context is drawn from the United States, Canada, the Eurozone, the United Kingdom, Switzerland and Japan. Each market is its own universe, and scores from different markets are not on one scale. StockLens does not score sectors, industries, ETFs, funds, indices, options or currencies, and does not cover India. Scores answer for the 350 largest companies by market value in each market with no account and on Free, the 6,000 largest worldwide on Pro and every company StockLens scores on Max, plus any company in your own portfolios, however deep it ranks. Every score carries the date it was computed.

Protocol transparency

What MCP standardizes, and what StockLens commits to

The Model Context Protocol is an open contract between an AI client and a tool server. StockLens implements it so any MCP-aware assistant -- ChatGPT, Claude, or a custom client -- can discover, call, and verify our analysis tools without bespoke glue code.

Together, these commitments define what each tool exposes, how calls are authorized, and which actions can change account state.

Typed tool registration

Tools declare their schema

Each StockLens tool publishes its name, input parameters, and output shape through MCP discovery. Clients see structured capabilities, not free-form prompts.

User-scoped auth

OAuth for account-scoped tools

Protected tools negotiate OAuth 2.0 Authorization Code with PKCE and run as the resolved StockLens user. Connectivity and the latest nightly market summary remain public; an account widens their reach.

Capability surface

Reads, one analysis trigger, and portfolio edits

Agents read pre-computed nightly scores, check analysis progress, and read any one of the authenticated user’s portfolios. Four tools change state: initiating an analysis may start private compute and consume a finite credit, two portfolio tools add or remove one company, and one adds a list the user approves after a preview. Nothing writes a holding quantity, a transaction or broker data, and no tool places a trade.

Safety boundaries

No PII, no cross-user access

Tool calls never carry personal identity -- handlers operate on your account and subscription tier alone, never your email or profile. Requests for analyses you cannot access return a generic not-found response, so the system never reveals another user’s activity.

Allowances

Research and analysis limits by plan

Each account has one weekly allowance of research requests, shared by ChatGPT, Claude and Codex, which renews every Monday at 00:00 UTC. Analysis tool calls renew every day at 00:00 UTC. Both are separate from web-app credits.

Free

50 research requests a week

10 standard analysis tool calls a day · No Plus Analysis

Pro

1,000 research requests a week

50 standard + 25 Plus Analysis tool calls a day

Max

Unlimited research requests, subject to fair use (up to 1,000 a day)

Unlimited analysis tool calls, subject to fair use

A research request is one successful company reading, shortlist reading of up to 10 companies, ranked screen or review of a portfolio in which at least one company is scored. Evidence inspection, report retrieval, adding or removing portfolio companies, retries and errors are free.

Readings without an account are limited to 20 a week for each network address, and to 2,500 a week (at most 800 a day) shared by everyone who reaches StockLens through an assistant that relays all its users through one published address.

Access pathways

How to point your AI at StockLens

Connect from the ChatGPT Plugin, the Claude Connector, or any MCP-compatible client. Three tools answer with no account at all, for the 350 largest companies in each market, so you can ask for a score before you connect anything. The other eight tools use OAuth 2.0 Authorization Code with PKCE.

Claude Connector

  1. Open Customize → Connectors in Claude, select Discover and search for StockLens. Team and Enterprise owners add it for the organization first
  2. Select Connect to Claude and sign in with your StockLens email via Google, Apple, or an emailed code
  3. Enable StockLens from the + menu in any chat

Read the guide for Personal Finance in Claude

Any MCP client

  1. Point your MCP-compatible client at https://mcp.stocklens.com/mcp
  2. Ask for a score straight away, before you sign in
  3. Sign in with your StockLens email via Google, Apple, or an emailed code when you first use a portfolio or analysis tool

ChatGPT Plugin

  1. Install StockLens from the ChatGPT plugin directory
  2. Ask for a score straight away, before you sign in
  3. Sign in with your StockLens email via Google, Apple, or an emailed code when you first use a portfolio or analysis tool

After you log in, Settings → AI Integrations has your endpoint and the exact setup steps for each client.

Explore More

Capabilities that work together

Multi-Domain Analysis How StockLens reads a company across technical, fundamental, risk, sentiment and macro, with the evidence for the reading and against it. Learn more

Portfolio Intelligence Overnight portfolio health monitoring, a weekly review, score change detection, and watchlist management. Learn more

Stock Discovery Choose a market, then rank its companies. Learn more

{"@context":"https://schema.org","@graph":[{"@type":"WebPage","name":"AI Integrations for ChatGPT & Claude | StockLens","description":"Where a company stands in its market, and why, in the ChatGPT Plugin, the Claude Connector or any MCP client. The 350 largest companies in each market answer with no account.","url":"https://stocklens.com/features/ai-integrations","isPartOf":{"@id":"https://stocklens.com/#website"},"about":{"@id":"https://stocklens.com/#software"},"publisher":{"@id":"https://stocklens.com/#organization"},"breadcrumb":{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://stocklens.com"},{"@type":"ListItem","position":2,"name":"AI Integrations","item":"https://stocklens.com/features/ai-integrations"}]}},{"@type":"WebSite","@id":"https://stocklens.com/#website","name":"StockLens","alternateName":"Stock Lens","url":"https://stocklens.com","publisher":{"@id":"https://stocklens.com/#organization"}},{"@type":"Organization","@id":"https://stocklens.com/#organization","name":"StockLens","legalName":"Stock Lens Inc.","url":"https://stocklens.com","description":"StockLens is a web-based stock research tool for self-directed investors. It scores listed companies within their own listing market, on company fundamentals alongside technical, risk, sentiment, and macro signals, and runs in a web browser at stocklens.com.","disambiguatingDescription":"StockLens is the web-based stock research tool published at stocklens.com by Stock Lens Inc. It is not affiliated with mobile iOS or Android chart-scanning apps, trading signal generators, or third-party GitHub repositories that share the StockLens name.","logo":{"@type":"ImageObject","url":"https://stocklens.com/logo512.png","width":512,"height":512},"sameAs":["https://x.com/stocklens_app","https://www.linkedin.com/company/stocklens","https://www.tiktok.com/@stocklens_app","https://www.instagram.com/stocklens_app/","https://www.facebook.com/people/StockLens/61568082762850/","https://www.reddit.com/r/stocklens/","https://www.crunchbase.com/organization/stocklens-e9fe"],"identifier":[{"@type":"PropertyValue","propertyID":"CIPO Canadian trademark registration","value":"TMA1435281"}],"contactPoint":{"@type":"ContactPoint","contactType":"customer support","email":"support@stocklens.com"},"address":{"@type":"PostalAddress","streetAddress":"PO Box 30025 RPO King St. W","addressLocality":"Toronto","addressRegion":"Ontario","postalCode":"M5V 0A3","addressCountry":"CA"},"founder":{"@type":"Person","@id":"https://stocklens.com/about#founder","name":"Clint Deygoo","url":"https://stocklens.com/about"}},{"@type":"SoftwareApplication","@id":"https://stocklens.com/#software","name":"StockLens","url":"https://stocklens.com","applicationCategory":"BusinessApplication","applicationSubCategory":"Stock Research","operatingSystem":"Web browser","disambiguatingDescription":"StockLens is the web-based stock research tool published at stocklens.com by Stock Lens Inc. It is not affiliated with mobile iOS or Android chart-scanning apps, trading signal generators, or third-party GitHub repositories that share the StockLens name.","developer":{"@id":"https://stocklens.com/#organization"},"offers":{"@type":"Offer","price":"0","priceCurrency":"CAD","url":"https://stocklens.com"}}]}