MCP.so
登录
D

DalalOS MCP Server

@DalalOS

基本信息

传输方式

streamable-http

认证方式

无需认证

最近更新

分类

其他

作者

DalalOS

关于 DalalOS MCP Server

Hosted MCP server for Indian (NSE/BSE) stock research. Gives Claude, ChatGPT, Cursor or any MCP client 59 tools: end-of-day quotes and price history, financials and ratios, peer comparison, shareholding and promoter pledge, corporate actions, filings with extracted facts, screens, indices, F&O and FII/DII context. Every response carries its source and freshness date. Sourced facts and mechanical measures, never buy/sell calls or target prices. Free tier, no card.

连接信息

https://mcp.dalalos.in/mcp

接入方式

claude mcp add dalalos --transport http https://mcp.dalalos.in/mcp

DalalOS MCP Server 工具

74

Use this only to confirm the DalalOS connection is working. Returns "pong". It does not check how current the data is; get_data_freshness does.

Use this when the user asks for the contents or a summary of one specific BSE filing of an Indian company, using a news_id from get_stock_events. It may queue processing and use a small daily allowance. Returns the facts extracted from the filing's PDF, each with a verbatim quote, and a summary when one is supported by those facts. `query` is the company's NSE symbol, ISIN or BSE code; `news_id` comes from get_stock_events or get_announcement_bucket. If the filing has not been extracted yet, the call queues it, waits up to about 30 seconds, and otherwise returns `status: "pending"` with `retry_after_seconds`: call again after that long. Queuing a filing counts against a small daily allowance; reading one that is already extracted does not. Financial-results filings are not extracted: use get_financials. For several filings of one company use get_filing_extracts. A response with no extraction carries a `note` saying why, and an `attachment_url` to the PDF where there is one.

Use this when the user wants summaries of several filings of one Indian company at once (up to 10 news_ids from get_stock_events). It may queue processing and use allowance. Returns one entry per distinct filing, in the order asked: the extracted facts with quotes, or a status such as `pending`, `not_queued`, `failed` or `skipped` with the reason. `news_ids` takes up to 10 filings of the same company. Filings not yet extracted are queued, each counting against the daily allowance, and the call never waits: it lists the ones it can be asked for again in `pending_news_ids` with one `retry_after_seconds`. Call again with just those. A large filing may come back as a stub listed in `omitted_for_size`; read it with get_filing_extract.

Use this when the user asks what FIIs and DIIs have been buying or selling in Indian equities day by day across the whole market. Not for a company's ownership percentage; use get_shareholding. Returns one row per trading day with FII and DII buy, sell and net values in Rs crore for the NSE cash segment. `from_date` and `to_date` are ISO dates; the default is the last 30 days and the longest window is 366 days. For F&O positions by participant use get_fo_participant_activity; for one sector, get_sector_fpi_investment.

Use this when the user asks about daily FII transaction flow in one Indian stock in the historical CDSL data, which stops at March 2025. Returns a daily series of FII/FPI buy, sell and net for one stock from CDSL's disclosures, with a 30-day summary. The data covers CDSL only and ends in March 2025, so it is a partial, historical view. `days` (1 to 1825, default 90) is the length of the series, counted back from the latest day of data held for the stock, not from today; `effective_from` and `effective_to` give the dates covered. For current flow use get_market_fii_dii_flow (whole market) or get_sector_fpi_investment (by sector); for holding percentages, get_shareholding.

Use this when the user asks what is on their DalalOS watchlist, a saved list of stock symbols (not brokerage holdings). Returns the saved symbols, newest first, each with name, sector and market cap where available. The list holds symbols only: no quantities, purchase prices or brokerage holdings. Needs a signed-in DalalOS account.

Use this when the user asks to add, watch or track Indian stocks in their DalalOS watchlist. Changes only the user's own saved list. `symbols` is a list or one string, and may be a pasted list separated by commas, semicolons, new lines, tabs or pipes (not spaces), up to 200 entries. Each entry may be an NSE symbol, ISIN, BSE code or company name, and is saved only when it identifies one NSE-listed company exactly. A misspelt or partial name is not saved: it comes back in `ambiguous` with the candidates, to confirm with the user. Returns `watchlist` (the entries saved by this call, including any already on the list), `unmatched` and `ambiguous`. Needs a signed-in DalalOS account.

Use this when the user asks to remove stocks from their DalalOS watchlist. Changes only the user's own saved list. `symbols` is a list or one string, and may be a pasted list separated by commas, semicolons, new lines, tabs or pipes, up to 200 entries. Each entry is a symbol as get_watchlist returns it, or the ISIN or BSE code of a company on the list; company names are not matched. To clear the list, call get_watchlist and pass every symbol it returns. Returns `removed` and `not_found`. Needs a signed-in DalalOS account.

Use this when the user asks to report a problem with DalalOS data or tools, or to request a feature, or agrees after you offer to report a likely DalalOS issue. Not for questions about a company; it sends a report to the DalalOS team and does not change any data. Saves the report and returns a report ID; it uses no credits. Required: `category`, `summary` (up to 160 characters), `details` (up to 2,000) and `request_id`, a new UUID for each report, reused only to retry the same report. Optional: `related_tool`, `security_identifier` with `security_identifier_type` (both or neither), `metric`, `observed_value` and `expected_value` (with units, up to 200 characters each), `as_of_date` and `source_reference`. Send only what the report needs: no conversation text, credentials, contact details or holdings. Say whether the problem is suspected or confirmed. A failure of this tool is not something to report with it. On success, give the user `data.message`, which carries the report ID; the report is saved, not yet reviewed, and the team does not reply to individual reports. Needs a signed-in DalalOS account.

Use this when the user names an Indian company, NSE symbol, BSE code or ISIN and you need to identify the listed security before calling another tool. Not for US or other non-Indian stocks. Returns matching securities with ISIN, NSE symbol, BSE code, name, sector and industry. It accepts part of a name, a former name, and everyday names such as SBI or L&T; `resolved_via` says when a match came from a former or everyday name. When nothing matches, the nearest names come back as suggestions with `reason_code: "typo_suggestion"`: confirm one with the user before using it. `limit` is 10 by default and at most 50, and a broad name can have more matches than are returned. Pass a returned NSE symbol, ISIN or BSE code as `query` to any per-company tool.

Use this when the user asks for the price, trading volume, 52-week range, PE or PB of an Indian company listed on NSE or BSE. Prices are end-of-day, not live. Not for non-Indian stocks or live intraday prices. Returns the latest close with open, high, low, volume, turnover, market cap, P/E, P/B, the 52-week high and low, and recent dividends. `query` is an NSE symbol, ISIN or BSE code. `as_of` is the trading day the quote is from. For several stocks use get_batch_quotes; for past prices, get_price_history; for P/E or P/B over time, get_valuation_history.

Use this when the user wants end-of-day prices or ratios for several Indian stocks at once (up to 20), for example to line up peers. Returns one item per symbol, each with its own `status`; the fields of get_stock_quote are under `quote`. `symbols` takes up to 20 NSE symbols, ISINs or BSE codes. A symbol that is not found is reported in its own item and does not fail the call. For a table comparing similar companies use get_peer_comparison.

Use this when the user asks how an Indian company listed on NSE or BSE has performed financially: revenue, profit, margins, balance sheet, cash flow or growth, quarterly or annual. Not for non-Indian companies; for a period-by-period growth series use get_financial_trends. Returns the profit and loss, balance sheet and cash flow for each period, consolidated where the company files consolidated results, with margins, the group's profit and the owners' share of it, a trailing-twelve-month block, and an annual growth summary (year-on-year and 3- and 5-year growth in revenue, net profit, EBITDA and EPS). `period_type` is "quarterly" (default) or "annual"; `limit` is the number of periods, newest first (default 8, at most 20). Quarterly results carry no cash flow and the balance sheet is filed half-yearly: use `period_type="annual"` for cash flow. The trailing-twelve-month block comes only with quarterly periods. Banks and NBFCs report interest income and net interest income, and insurers premium income, in place of revenue and EBITDA. Trust units return filed trust results: distribution per unit, NAV and cash flow. For growth and margins period by period, or segments over time, use get_financial_trends.

Use this when the user asks what an Indian company listed on NSE or BSE is: its sector, industry, market cap and a headline financial summary. Returns the company's identity and classification, its latest end-of-day market figures (price, market cap, P/E, P/B, 52-week range), a summary of the latest quarter and the trailing twelve months, and, where its annual report has been read, a description of the business and its management. For full statements use get_financials; for a sourced report across topics, get_company_report.

Use this when the user asks a broad question about an Indian company listed on NSE or BSE ("how is X doing", "should I look at X"): a sourced report on quarterly and annual results, valuation, ownership, pledge, dividends and recent disclosures, no verdict. For one topic use get_financials, get_shareholding or get_stock_events. Returns one report built from DalalOS's stored data: facts with their sources, calculations, the statements those support, recent events, and a `coverage` block naming what is missing. It draws on results, valuation history, returns, ownership and pledge, dividends, corporate actions, peers and recent disclosures. `query` is an NSE symbol, ISIN or BSE code; `days` (1 to 365, default 90) sets how far back disclosures are included. The default is a compact report: the latest three quarters and two years, material events in full and routine filings as a line each. `sections` returns the named sections in full, and `detail="full"` returns everything, which is too large for most clients to read.

Use this when the user compares an Indian company listed on NSE or BSE with similar companies on valuation, margins, return on capital or debt (PE, PB, ROE, debt-to-equity). For peers chosen by business, with reasons, use get_relevant_peers. Returns a table of up to `limit` companies (default 8, at most 15), the company asked about among them, from the same SEBI industry and of similar size (between one-tenth and ten times its market cap, largest first), with the medians across the table. Columns, on a trailing-twelve-month basis: market cap, P/E, P/B, revenue, net and EBITDA margin, ROE, ROA, ROCE, interest coverage, debt to equity, enterprise-value multiples and cash-flow ratios. Peers are chosen by classification and size, not by how alike the businesses are: for peers with reasons use get_relevant_peers. To compare companies the user names, call get_financials or get_batch_quotes with those companies.

Use this when the user asks who owns an Indian company listed on NSE or BSE: promoter, FII, DII, mutual fund and public holding percentages by quarter. Not for daily buying or selling; for daily FII/DII flow use get_market_fii_dii_flow or get_stock_fii_flow. Returns the shareholding pattern for each quarter, newest first: promoter, FII, DII and non-promoter percentages with the change from the quarter before, the promoters' pledged shares as a percentage of their own holding and of all shares, and, where available, institutions split by type and the holders named in the filing. `limit` is the number of quarters (default 8, at most 20). For the pledge series alone use get_pledge_trend.

Use this when the user wants past share prices of an Indian company listed on NSE or BSE over a period, daily, weekly or monthly, for example to chart or compare a move. Returns end-of-day bars (open, high, low, close, volume) with 50- and 200-day averages. `days` (1 to 1825, default 90) counts calendar days back from today. `interval` is "daily", "weekly" or "monthly"; a daily series longer than 750 bars comes back weekly. `adjusted=true` rescales prices for splits and bonuses; the default is prices as traded. For returns over standard periods use get_returns; for bonds, funds and SME listings, get_security_history.

Use this when the user asks how an Indian stock's PE, PB (price-to-book, banks included), price-to-sales or market cap has changed over time. Returns a series of P/E, P/B, price to sales, market cap and net debt, and for non-financial companies the enterprise-value multiples. Each point uses that day's close and only the results that were public by then. `days` (1 to 1825, default 365) counts back from today; `interval` is "daily", "weekly" or "monthly". P/B is given for banks, NBFCs and insurers; their enterprise-value fields are empty.

Use this when the user asks how much an Indian stock has returned over a period (1 week, 1, 3 or 5 years, or its full history), or about its volatility or drawdown. Returns price returns for 1 week, 1, 3 and 6 months, 1, 3 and 5 years, the year to date and the full history, with annualised returns, volatility, the largest fall from a peak, and the 52-week high and low. Returns use prices adjusted for splits and bonuses and do not include dividends. There is no period parameter: for another period use get_price_history or get_stock_vs_index.

Use this when the user asks for moving averages, RSI, MACD, Bollinger Bands, beta or circuit-limit data for an Indian stock. Returns the latest values of the moving averages, RSI (14), MACD (12, 26, 9), Bollinger Bands (20, 2) and beta against the Nifty 50, with the stock's circuit-filter band and the days it closed at that band. Latest values only: no history of an indicator and no signals. An indicator without enough history is empty and named in `insufficient_history`.

Use this when the user asks how much of an Indian stock's traded volume was taken for delivery, and how that has trended. Returns, for each day, the shares traded, the shares taken for delivery and the delivery percentage, with the average over the window and over the last five sessions. `days` (1 to 1825, default 90) counts back from today; a long daily series comes back weekly.

Use this when the user asks whether and when an Indian company paid or declared a dividend, or about its dividend yield. Returns each cash dividend with its ex-date, type and amount per share, the total for the last twelve months with the yield and payout ratio on it, and totals by financial year (April to March). `years` (1 to 15, default 5) sets how far back. The list includes dividends announced with an ex-date still to come. For splits and bonuses use get_corporate_actions.

Use this when the user asks about an Indian company's BRSR (Business Responsibility and Sustainability Report) or ESG disclosures. Returns the figures from the company's Business Responsibility and Sustainability Report as filed, by financial year, newest first: emissions, energy, water, waste, workforce and governance. `years` is 1 to 10 (default 3). Only the 1,000 largest listed companies file this report. A figure that is absent was not in the filing, or is not one DalalOS reads; it is not zero. Leave `include_raw` unset: the full filing is not returned here.

Use this when the user asks about bulk or block trades in an Indian stock on the NSE. Returns the bulk and block deals NSE disclosed in the stock, newest first: date, client, buy or sell, quantity and price. Each row is one client on one side; NSE does not pair a buyer with a seller. `deal_type` is "bulk" or "block" (default both); `days` is 1 to 1825 (default 90); `limit` is 1 to 500 (default 100). Deals are held from the day DalalOS began collecting them, so older ones may be absent. For trades by promoters and directors use get_insider_trades.

Use this when the user asks which promoters, directors or company insiders bought, sold or pledged shares of Indian listed companies, market-wide or for one company. Not for bulk or block deals by funds; use get_bulk_block_deals. Returns disclosed trades and pledges by promoters, directors, key managers and designated persons, from BSE's insider-trading disclosures. Omit `symbol` for the whole market, or pass an NSE symbol, ISIN or BSE code. The window is by trade date: `days` back from today (1 to 365, default 30), or `from_date` and `to_date` (ISO, at most 365 days apart). `category` is "promoters" (default), "directors_kmp", "designated", "relatives" or "all". `side` is "trades" (buys and sells, default), "buy", "sell", "pledge" (pledges and releases) or "all". `min_value_inr` keeps rows of at least that value and drops rows with no disclosed value. `group_by` is "trade" (default, newest first) or "company" (buy, sell and net value per company, ordered by total value traded). `limit` is 1 to 500 (default 100). Equity only unless `include_derivatives` is true. Reports for the latest two weeks are still arriving.

Use this when the user asks about open interest, put-call ratio or futures and options positioning for an Indian stock or an index such as NIFTY or BANKNIFTY. Returns the latest trading day's totals across all contracts of one F&O underlying: futures open interest and its change, call and put open interest, the put-call ratio (put open interest divided by call open interest) volume, and whether it is in the ban list. `query` is a stock's NSE symbol, ISIN or BSE code, or an index name such as NIFTY or BANKNIFTY. Only stocks that trade in F&O have data. For positions by participant type across the market use get_fo_participant_activity.

Use this when the user asks how FIIs, DIIs, proprietary traders or clients are positioned in NSE futures and options across the whole market. Returns one row per day and participant type (FII, DII, proprietary, client) with long and short contract counts in index and stock futures and options, across the whole F&O segment: open interest, which is positions held, and the day's volume. `days` is 1 to 366 (default 10). Neither is cash-market buying and selling, which is in get_market_fii_dii_flow. For one stock or index use get_derivatives_positioning.

Use this when the user asks which Indian stocks are in the NSE futures-and-options ban period today. Returns the securities in the F&O ban period on the latest day NSE published the list, with that date in `ban_date`. It takes no arguments. An empty list means nothing is in ban that day. To check one stock together with its open interest use get_derivatives_positioning.

Use this when the user asks about an Indian company's splits, bonuses, dividends, buybacks or other corporate actions, past or announced. For events across the whole market use get_market_calendar. Returns the company's corporate actions, newest ex-date first, including any announced with an ex-date in the coming weeks. `years` is 1 to 20 (default 5). `types` filters to a list drawn from: dividend, bonus, split, consolidation, rights, buyback, demerger, spinoff, merger, distribution, other. For dividends with yield use get_dividends; for events across the market, get_market_calendar.

Use this when the user asks for the latest news on an Indian company, or what it has announced or filed recently: orders, board meetings, fundraising, acquisitions, management changes. It lists the company's own BSE disclosures, not media news. Returns the company's BSE announcements, newest first: a readable `title`, category, filing time, a link to the PDF and a `news_id`. It lists filings and does not read them: pass `news_id` values to get_filing_extracts for their contents. `days` is 1 to 365 (default 30). `category` filters to one of: Company Update, Board Meeting, Corp. Action, AGM/EGM, Insider Trading / SAST, Integrated Filing, Others; any other value returns nothing. Announcements that attach the same document are shown once. The results announcement itself is not listed: the figures are in get_financials. A long list is cut to fit, with `truncated` set. Listing may queue a few of the filings for reading, which counts against the daily filing allowance. It also lists the company's index changes.

Use this when the user asks what a term or ratio on an Indian company's page means or how it is calculated: P/E, ROE, TTM, gross NPA, promoter pledge and the like. It defines terms and never says whether a value is good or bad. Returns plain-word definitions of the terms, line items and ratios DalalOS reports: what each is, how it is calculated, its unit and which kinds of company it applies to. `term` is one or more keys separated by commas, such as `pe,market_cap`; one unknown key fails the call. `kind` filters to "metric", "line_item", "term" or "section". Without `term` it returns only the list of keys, so pass `term` for the definitions you want.

Use this when the user wants the documents an Indian company filed, grouped by type: results, annual reports, investor presentations, call transcripts and recordings, board meeting and AGM papers, ratings. For what was announced and when use get_stock_events. Returns the documents the company filed with BSE, each with its type, date and link. `days` is 1 to 1825 (default 365). `document_type` filters to one of: results, annual_report, presentation, transcript, recording, board_meeting, agm_egm, insider_takeover, rating, management_change, deal, corporate_action, press_release, investor_meet, compliance, other. `limit` is 1 to 1000 (default 60), newest first: pass a type or fewer `days` when the list is cut. Transcripts and recordings are links only. The `results` type can be incomplete: the figures are in get_financials. A document with an `extraction_id` has already been read: pass that id as `news_id` to get_filing_extract.

Use this when the user needs historical detail from an Indian company's earlier filings, for example a prior order value or commissioning date. Returns passages from the company's filings that contain the search words, each with its page number, filing date and a link to the PDF. `search_text` is a few distinctive words: a passage matches if it contains any of them, best matches first, and words such as AND or OR are not operators. `from_date` and `to_date` (ISO) limit the filing date. `document_types` limits the kind of filing to a list from: concall, presentation, order_win, rating, debt_security, acquisition, results, press_release, sast_disclosure, corp_action, governance_change, fund_raise, scheme_of_arrangement, business_update, annual_report, brsr, meeting_notice, compliance_noise, other; any other value matches nothing. `limit` is at most 20 (default 10), with at most two passages per filing. Only filings DalalOS has already read are searched, so no match does not mean the company never disclosed it; `coverage` shows what is indexed. This tool never queues a filing for reading.

Use this after get_filing_extract when the user wants the earlier state behind an order, capacity or similar event in a filing. Returns, for one filing given by `news_id` (from get_stock_events), the event recorded from it, the earlier event it follows, what changed between them, related disclosures, the timeline of that group of filings, and the questions those filings leave open. It covers events such as orders, capacity, acquisitions, guidance, credit ratings, shareholding changes and resignations. The filing must have been read first with get_filing_extract. Only links that the filings themselves establish are reported, so an empty link means none was established, not that there is no history.

Use this when the user asks for an Indian company's disclosed operating figures such as order book, order inflow, capacity or other business KPIs over time. Returns the operating figures the company disclosed in its filings, each with its date, unit and source, and a `coverage` entry per metric saying whether it was found. `metrics` is a list from: order_book, order_inflow, execution_period, working_capital_days_change, capacity, installed_capacity, planned_capacity, power_capacity, utilisation, commissioning_date, announced_capex, capex_spent, remaining_capex, store_count, inventory, aum, disbursements, net_interest_margin, casa_ratio, average_interest_earning_assets; the default is all of them. `from_date` and `to_date` (ISO) limit the dates and `latest_only=true` gives the newest value of each. `limit` is at most 100 and is applied across metrics in alphabetical order, so name the metrics you need. Values are as disclosed: an order book is never derived from new orders.

Use this when the user asks how an Indian company is exposed to a macro factor such as crude oil or the rupee-dollar rate, and whether the latest official move in it is large enough to matter. Returns the company's reviewed exposures to macro factors that apply on `as_of_date` (ISO, default today), and for crude oil and USD/INR whether the latest official move (PPAC Indian crude basket, FBIL reference rate) is large enough to flag, with a `coverage` entry for every factor asked about. Only crude oil and USD/INR are monitored; other factors report that no source is available. `factors` is a list from: crude_oil, usd_inr, interest_rates, gold_price, silver_price, steel_price, aluminium_price, government_capex, real_estate_activity, power_sector_capex, renewable_policy_and_tenders. It gives context and never says a move caused or will cause a result. For the plain list of exposures use get_company_exposures.

Use this when the user asks which macro or commodity factors an Indian company depends on, based on reviewed exposure mappings. Returns every reviewed exposure of the company to a macro or commodity factor: the factor, how it reaches the company (direct or indirect, direction, lag), any stated sensitivity and mitigants, the dates it applies, and the source quote. Entries are not filtered by date, so check `valid_from` and `valid_to`. A company with none reviewed returns an empty list. To see whether a current move in crude oil or USD/INR is large enough to matter use get_company_macro_exposures.

Use this when the user asks who an Indian company's real peers or competitors are and why. For a valuation table of peers use get_peer_comparison. Returns up to `limit` peers (default 8, at most 15) chosen from reviewed descriptions of what each business does, each with a similarity score, what is shared, what differs, and figures to compare. `purpose` is "operating" (default), "financial" or "valuation". It can return few peers or none when the evidence is thin: it does not fall back to companies that only share an industry label. For a table of ratios across the industry use get_peer_comparison.

Use this when the user wants to browse market-wide disclosure types, such as order receipts or fundraises, before calling get_announcement_bucket. Returns the kinds of disclosure that can be listed across the market: order-receipts, fundraise, credit-rating, acquisitions and management-changes. It takes no arguments. Pass one to get_announcement_bucket.

Use this when the user asks which Indian companies recently announced a given kind of disclosure, for example order wins or fundraises, across the whole market. Returns BSE filings of one kind across all companies, newest first, each with the company, headline, filing time, a link to the PDF and a `news_id`. `bucket` is one of: order-receipts, fundraise, credit-rating, acquisitions, management-changes. `days` is 1 to 90 (default 7). To read a filing, pass its `news_id` and that row's `nse_symbol` or `isin` to get_filing_extract. A long list is cut to fit, with `truncated` set. Listing may queue a few of the filings for reading, which counts against the daily filing allowance. For one company's filings use get_stock_events.

Use this when the user asks for Indian stocks matching numeric criteria such as PE, PB, market cap, ROE, margins or debt, optionally within an index such as NIFTY 50. For OR logic or arithmetic between metrics use screen_by_query. Returns up to `limit` securities (default 25, at most 100) that meet every filter, sorted by `sort_by` (default market_cap) in `order` ("desc" or "asc"), with the filtered metrics on each row. `filters` is an object. A range filter is a metric name ending in `_min` or `_max`, and the unit depends on the metric. Fractions, where 0.15 means 15%: roe, roa, roce, net_margin, ebitda_margin, operating_margin, dividend_yield, earnings_yield, fcf_yield, revenue_growth, net_income_growth, return_1d to return_1y, the 3- and 5-year growth rates, pct_off_52w_low. Percentages, where 50 means 50%: promoter_pct, pledged_pct, deliv_pct, and the quarterly change in promoter, FII, DII and public holding, in percentage points. Rupees: market_cap, ev, price, turnover, week52_high, week52_low (Rs 5,000 crore is 50000000000). Plain numbers: pe, pb, debt_to_equity, interest_coverage, ev_to_ebitda, price_to_sales, price_to_cash_flow, piotroski_f_score (0 to 9). Exact filters take one value: sector, industry, igroup, isubgroup, company_type (general, bank, nbfc, insurer, reit, invit), index (an index name such as "NIFTY 50") and is_sme (true or false). A name that is not recognised returns the closest valid names in place of an empty result. More metrics are accepted than are listed here: an unknown key returns an error that lists them all. A security with no value for a filtered metric is left out and counted in `excluded_missing_metric`; margin, debt and enterprise-value metrics have no value for banks, NBFCs and insurers.

Use this when the user's screen needs OR logic, arithmetic between two metrics or a screen expression that screen_stocks filters cannot express. Returns matching securities, 25 a page by default, with identity and core valuation columns and the metrics the query names; name any other metric you want returned in `columns`. `query` is one expression of metric names, numbers, = != > < >= <=, AND, OR, brackets, + - * / and the functions COALESCE, IF, LEAST, POWER, ABS, SQRT and LOG, for example `Return on capital employed > 22% AND (Debt to equity < 0.5 OR Interest Coverage Ratio > 8)`. It has no text values: restrict by sector, industry, igroup, isubgroup, company_type or index with `scope`, an object such as {"index": "NIFTY 50"}; a name that is not recognised is refused with the closest valid names. `dialect` is "screener" (default: common screener names, percentages written as 22%, market cap in crore) or "dalalos" (DalalOS metric ids in stored units: fractions and rupees). An unknown metric name is an error, never a guess. A comparison with a missing value does not match; wrap the metric in COALESCE to give it a default. `limit` is 1 to 100 and `offset` at most 5000. Pass `pagination.snapshot_id` from the first page as `snapshot_id` on later pages, and start again from `offset` 0 if a call returns `snapshot_expired`. `definitions` adds named ratios to use in the query.

Use this when the user asks how an Indian sector or industry looks in aggregate: number of companies, market cap, valuation and returns. Returns, for each group, the number of companies, total market cap, median and average P/E and median P/B, and where available the median dividend yield and one-year return. `grain` is "sector" (default), "igroup" (SEBI industry) or "isubgroup" (SEBI basic industry). Omit `sector` to list every group; pass a name at that grain for one group, and a name that is not recognised returns the valid ones. Totals only: no company rows.

Use this when the user asks which stocks are in an Indian index such as NIFTY 50 or NIFTY Bank. Returns the stocks in an index, largest first, with ISIN, NSE symbol, sector and market cap. `index` is an index name such as NIFTY 50 or BSE SENSEX; a name that is not recognised returns the valid ones. `limit` is 1 to 500 (default 100). It gives membership only, not index weights.

Use this when the user needs the list of Indian market indices DalalOS covers. Returns every index whose membership DalalOS holds, with the number of members. These names are the `index` values for list_index_constituents, get_market_heatmap, get_top_movers and screen_stocks. For indices with P/E and P/B history use list_index_valuations.

Use this when the user wants valuation and returns across the stocks in an Indian index at a glance. Not for ranking by return; use get_top_movers. Returns each stock in an index with market cap, P/E, P/B and its return over 1 day, 1 week and 1, 3, 6 and 12 months, largest company first. `index` is an index name such as NIFTY 50; a name that is not recognised returns the valid ones. `limit` is 1 to 500 (default 100). The order is by size and cannot be changed: to rank by the day's move use get_top_movers.

Use this when the user asks which Indian stocks in an index gained or lost the most, or traded the most, in the latest end-of-day session. Returns the stocks in an index ranked on their latest end-of-day quote. `metric` is "pct_change" (the day's percentage change, default), "change" (in rupees), "turnover" (rupees traded) or "volume" (shares traded). `direction` is "desc" (default: largest first) or "asc". `index` defaults to NIFTY 500; a name that is not recognised returns the valid ones. `limit` is 1 to 100 (default 20). The session cannot be chosen: the result's `as_of` says which one was ranked. To rank every NSE security on a past day use get_market_ranking.

Use this when the user asks what ready-made screens are available, before calling run_preset_screen. Returns the ready-made screens, each with a description and the time its result was last computed: low_price_to_earnings, high_dividend_yield_large_market_cap, low_leverage_high_roe, high_promoter_holding_low_pledge and high_growth_latest_results. It takes no arguments. Run one with run_preset_screen; for your own conditions use screen_stocks.

Use this when the user wants the latest result of a named ready-made screen from list_preset_screens. Returns the stored result of one ready-made screen, computed nightly: up to 50 rows, the filters it applied and the number of matches. `name` is one of the names from list_preset_screens. For more rows or other thresholds, pass the returned `filters` to screen_stocks.

Use this when the user asks which Indian companies report results soon, optionally by sector. For a past window or other event types use get_market_calendar. Returns companies with a board meeting for results in the next `days` days (1 to 90, default 14), earliest first, with the date and identifiers. `sector` filters to one sector; a name that is not recognised returns the closest valid names, and companies DalalOS has not matched to a sector are left out. `limit` is 1 to 200 (default 50); `truncated` says when later dates were cut. For past dates or other kinds of event use get_market_calendar.

Use this when the user asks for an Indian index's PE, PB or dividend yield over time, for example whether NIFTY 50 is expensive relative to its history. Returns one index's daily open, high, low, close, P/E, P/B and dividend yield as NSE publishes them. `index` is an NSE index name such as Nifty 50, Nifty Bank or Nifty IT; the names are in list_index_valuations. `from_date` and `to_date` are ISO dates at most 1825 days apart; the default is the 30 days to today. A long window returns the most recent rows. An empty result can mean the window holds no trading days, not that the name is wrong. For return and volatility across indices use compare_indices.

Use this when the user needs the names of the NSE indices that have daily close, PE, PB and dividend-yield history. Returns every NSE index DalalOS holds daily data for, with its latest date and close. These are the names get_index_valuation, compare_indices and get_stock_vs_index accept. It lists names, not valuations.

Use this when the user asks about the India VIX, the NSE volatility index, today or over a period. Returns the daily open, high, low, close and change of the India VIX. `from_date` and `to_date` are ISO dates at most 1825 days apart; the default is the 30 days to today.

Use this when the user asks which Indian stocks had the most volume, turnover, trades, delivery percentage or the largest move on a given trading day. For today's movers within one index use get_top_movers. Returns NSE securities ranked on one trading day. `metric` is "volume" (default), "turnover", "num_trades", "deliv_pct" or "pct_change"; `direction` is "desc" (default) or "asc"; `limit` is 1 to 100 (default 25). `date` is an ISO date, by default the latest day held. `instrument_class` is "equity" (default), "sme", "fund_unit" or "all". `series` narrows to one NSE series such as EQ or BE and applies together with `instrument_class`: for an SME or bond series pass `instrument_class="all"`. A security with no figure for the metric is left out. For movers within one index on the latest session use get_top_movers.

Use this when the user wants the raw daily NSE rows for a security, including SME stocks, bonds, G-secs, SGBs or fund units, with delivery and trade counts. Returns the daily rows NSE published for one security in every series it trades in, newest first: open, high, low, close, volume, turnover, trades and delivery. It covers SME stocks, bonds, government securities, gold bonds and fund units, which get_price_history does not. `symbol` is an NSE symbol or ISIN; `series` narrows to one, such as EQ, BE, SM or GB. The window is `days` back from the latest day held (default 30, at most 1825), or `from_date` and `to_date`. Prices are as published and not adjusted for splits or bonuses: for an adjusted series use get_price_history.

Use this when the user asks for NSE's own market cap and shares in issue, for one company over time or the largest companies on a day. Returns NSE's published market capitalisation, shares in issue and face value. With `symbol` (an NSE symbol): that security day by day over `days` (default 30, at most 1825) or `from_date` and `to_date`; `date` and `limit` are then ignored. Without `symbol`: the `limit` largest companies (1 to 100, default 25) on `date` (default the latest day held), with NSE's totals for the market; the window arguments are then ignored. NSE publishes it from 1 February 2024; stored depth depends on seeding.

Use this when the user asks which Indian stocks hit their upper or lower circuit on a day, or how often a stock has. Returns NSE's record of securities that hit their upper or lower price band. With `date` (default the latest day held): the hits on that day. With `symbol`: that security's hits over `days` (default 90, at most 1825), newest first. Pass one or the other, not both. `direction` is "upper", "lower" or "all" (default). `limit` is 1 to 500 (default 200), and `truncated` says when there are more. This is the exchange's own list: a stock can touch its band during the day and close inside it, so closing prices do not reproduce it. NSE publishes records from 2010; stored depth depends on seeding.

Use this when the user wants several Indian indices compared on return, volatility and drawdown over the same period. Returns, for each index over the same window, the price return, annualised volatility, largest fall from a peak, and best and worst day, computed from NSE's daily closes. `indices` is a list of 2 to 10 NSE index names, such as ["Nifty 50", "Nifty Bank"]; a name that is not found is reported along with the valid names. The window is `days` back from the latest day held (default 365, at most 1825), or `from_date` and `to_date`. Returns exclude dividends. For one index's P/E and P/B use get_index_valuation.

Use this when the user asks how an Indian stock's return compared with an index such as Nifty 50 over a period. Returns the stock's price return beside the index's over the same window, with each side's annualised volatility, largest fall from a peak, and best and worst day, and the difference between the two returns in percentage points. `symbol` is the NSE symbol of a mainboard equity; SME stocks, funds and bonds are not covered. `index` is an NSE index name (default Nifty 50). The window is `days` (default 365, at most 1825), or `from_date` and `to_date`. Returns exclude dividends and adjust for recorded splits and bonuses. `days_without_row` counts stored days the stock has no row.

Use this when the user asks how many NSE-listed securities advanced, declined or were unchanged on a trading day. Returns one row per trading day with the number of NSE securities that advanced, declined and were unchanged. `from_date` and `to_date` are ISO dates at most 366 days apart; the default is the 30 days to today. By default the counts are NSE's own published figures, held from the day DalalOS began collecting them. `universe` switches to counts computed from the daily price file, which go back further: "all_series" (every listed security, including funds and bonds), "equity_sme" (shares and SME shares) or "equity" (mainboard shares only). Use "equity" when the question is about stocks.

Use this when the user asks for official Indian macro data: CPI inflation, WPI, IIP, GDP or the crude-oil basket price. Returns official Indian macro series, newest period first. `dataset` is required: "cpi" (consumer prices, headline and by group), "wpi" (wholesale prices, overall and food only), "iip" (industrial production by category), "gdp" (annual level and growth, one row per published estimate) or "crude_basket" (the Indian crude basket monthly average, USD per barrel). `from_period` and `to_period` are YYYY-MM, or a fiscal year such as 2024-25 for gdp. `series_contains` keeps series whose name contains the text, such as Food or Manufacturing. `limit` is 1 to 2000 (default 200). For rupee exchange rates use get_fx_reference_rates.

Use this when the user asks for official FBIL reference rates of the rupee against USD, GBP, EUR, JPY, AED or IDR, for example the USD to INR rate. Returns FBIL's daily reference rates for the rupee against USD, GBP, EUR, JPY, AED and IDR. `currency_pair` filters to one pair, such as USD/INR. `from_date` and `to_date` are ISO dates; the default is the last 30 days. `rate` is rupees per `unit_base` units of the foreign currency: 1 for most, 100 for JPY, 10,000 for IDR. These are reference rates published on business days, not live quotes.

Use this when the user asks about daily FII flow by sector in the historical CDSL data, which stops at March 2025. For current sector flows use get_sector_fpi_investment. Returns daily FII/FPI buy, sell and net value by sector from CDSL's disclosures. The data covers CDSL only and ends in March 2025. `from_date` and `to_date` are ISO dates at most 366 days apart; when they are omitted the window is the 30 days up to the latest data held, and `effective_from` and `effective_to` give the dates covered. `sector` filters to one sector; a name that is not found returns the valid names. For current flow by sector use get_sector_fpi_investment.

Use this when the user asks which Indian sectors foreign portfolio investors are buying or selling, using fortnightly NSDL data. Returns, for each fortnight NSDL reports, FPI net investment and assets under custody by sector, in Rs crore and USD million, covering all FPIs across both depositories from 2012. `from_date` and `to_date` are ISO dates matched to the fortnight's end; the default is the last six months, and the longest window is five years unless `sector` is given. `sector` is one of NSDL's own sector names, or "Grand Total" for all sectors: call without `sector` to see the names. They changed on 30 April 2022, so a name may match only one side of that date. `instrument` selects the column: "equity" (default), debt_general, debt_vrr, debt_far, hybrid, aif, total, mf_equity, mf_debt, mf_hybrid, mf_solution_oriented or mf_other; before 2020 only equity, debt_general and total were reported. `include_subsectors` adds the older sub-sector rows, which are already counted in their parent.

Deprecated alias of get_stock_fii_flow; use get_stock_fii_flow for per-stock daily FII transaction flow. Returns the same result as get_stock_fii_flow, which replaces it: a daily series of FII/FPI buy, sell and net for one stock from CDSL's disclosures, ending in March 2025, with a 30-day summary. `days` (1 to 1825, default 90) is counted back from the latest day of data held, not from today.

Use this when the user asks what events are scheduled or happened across Indian markets in any date window: results, board meetings, dividends, splits. Returns disclosed events across all companies in a date window, past or future: board meetings for results, corporate actions, and IPO opening, closing and listing dates. `date_from` defaults to today and `date_to` to 30 days later; the window is at most 180 days. `types` filters to a list from: results, ipo, dividend, bonus, split, consolidation, rights, buyback, demerger, spinoff, merger, distribution, other. `limit` is 1 to 2000 (default 150). A list that was cut says so: ask again with `date_from` set to the last date returned, or with fewer `types`. An event's date is the board-meeting date for results, the ex-date for a corporate action, and the opening, closing or listing date for an IPO. For one company's actions use get_corporate_actions.

Use this when the user asks about Indian IPOs: upcoming, open, awaiting listing or recently listed. Returns public issues with price band, issue size, opening, closing and listing dates, the closing price on listing day and a link to the prospectus, each with a `status`: "upcoming", "open", "listing_pending" or "listed". `status` filters to one of these; `limit` is 1 to 500 (default 100). Rows are ordered by opening date, latest first, whatever their status. Lot size, final issue price, subscription multiples and listing gain are not available, and there is no grey-market premium: DalalOS holds only what the exchanges and SEBI publish.

Use this when the user asks for quarter-on-quarter or year-on-year growth and margin trends of an Indian company period by period. Returns one row per period, newest first, with revenue, net profit, EPS and margins, the change from the period before and from the same period a year earlier, and segment figures, plus growth rates across the periods shown. `period_type` is "quarterly" (default) or "annual"; `limit` is 1 to 20 (default 8). A quarter the company reported only as a year-to-date figure is left out, and a change is empty when the period it compares with is missing. It has no balance sheet and no trailing-twelve-month figures: use get_financials for those.

Use this when the user asks how much of an Indian company's promoter holding is pledged and how that has changed. Returns, for each quarter, newest first, the percentage of the promoters' own shares that are pledged or otherwise encumbered, the same shares as a percentage of all shares, their number and the promoters' holding, with the latest figure and its change from the quarter before. `limit` is 1 to 16 (default 12). `pledge_disclosed` is false where the filing gave no pledge figure, which is not the same as zero. For the full ownership breakdown use get_shareholding.

Use this when the user asks how recent DalalOS data is, or when a figure looks stale. Returns, for each of DalalOS's refresh jobs, when it last succeeded and whether it is overdue, and for each dataset any failed quality check. It takes no arguments. It reports whether refreshes are running, not the date of the data: for one figure, read `as_of` and `cached_at` in that tool's own result. The envelope's `stale` is true whenever the overall status is not ok (a job overdue or stuck, or a failed quality check), so it does not mean every figure is out of date; `data.stale` lists the overdue jobs.

Use this when the user asks which stocks joined or left an Indian index or list (MSCI India, Nifty, Sensex, the F&O list, AMFI categories) and when, or which changes are announced but not yet effective. Returns the changes, newest first. Give `index` (id or name), `query` (NSE symbol, ISIN or BSE code), or `from_date` and `to_date` (YYYY-MM-DD, on the effective date). `change_type` is add, delete, segment_up, segment_down or weight_change. `limit` is 1 to 200 (default 100); `include_superseded` adds superseded and cancelled rows. For members now, or on a past date with `as_of`, use list_index_constituents.

Use this when the user asks how a stock traded around the announcement and effective dates of its entry to or exit from an index or the F&O list, alone or as a group. These are index dates, not results announcements. Returns stock and benchmark returns in windows around those dates, volume and delivery against a baseline, and flags for other events nearby. Give `index`; add `query` (NSE symbol, ISIN or BSE code) for one company, or leave it out for a group (statistics from 5 analysed events). `change_type` defaults to add; `from_date` and `to_date` bound the effective date; `benchmark` is an NSE index name (default Nifty 500); `include_flagged` false drops flagged events; `limit` is 1 to 200 (default 50).

Use this when the user asks how much index funds may need to buy of a stock joining an index, or its estimated weight there. Estimates with their inputs, not a price forecast or FII flows. Returns the estimated weight and the rupees, share of free float and days of median traded value it implies for money tracking the index, each input with its date and source; `not_computed` names what is missing. `query` is an NSE symbol, ISIN or BSE code; `index` an id or name. `tracking_assets` replaces the stored figure, in full units of `tracking_assets_currency` (default INR). For daily FII and DII flows use get_market_fii_dii_flow; for past events, get_index_event_study.

概览

What it is

Hosted MCP server for Indian (NSE/BSE) stock research. Gives Claude, ChatGPT, Cursor or any MCP client 59 tools: end-of-day quotes and price history, financials and ratios, peer comparison, shareholding and promoter pledge, corporate actions, filings with extracted facts, screens, indices, F&O and FII/DII context. Every response carries its source and freshness date. Sourced facts and mechanical measures, never buy/sell calls or target prices. Free tier, no card.

From the publisher's website: DalalOS is an MCP plug-in for Claude, ChatGPT and other AI apps. It gives them Indian company results, shareholding, FII and DII flows and IPO data.

Connect

  • Endpoint: https://mcp.dalalos.in/mcp (Streamable HTTP)
  • Authentication: no authentication

Claude Code:

claude mcp add dalalos --transport http https://mcp.dalalos.in/mcp

Other clients (mcpServers config):

{ "mcpServers": { "dalalos": { "type": "http", "url": "https://mcp.dalalos.in/mcp" } } }

Tools

Read from the server's tools/list on 2026-11-13:

  • ping — Use this only to confirm the DalalOS connection is working.

Returns "pong"

  • get_filing_extract — Use this when the user asks for the contents or a summary of one specific BSE filing of an Indian company, using a news_id from get_stock_events
  • get_filing_extracts — Use this when the user wants summaries of several filings of one Indian company at once (up to 10 news_ids from get_stock_events)
  • get_market_fii_dii_flow — Use this when the user asks what FIIs and DIIs have been buying or selling in Indian equities day by day across the whole market
  • get_stock_fii_flow — Use this when the user asks about daily FII transaction flow in one Indian stock in the historical CDSL data, which stops at March 2025.

Returns a daily series of FII/FPI buy, sell and net for one stock from CDSL's disclosures, with a 30-day summary

  • get_watchlist — Use this when the user asks what is on their DalalOS watchlist, a saved list of stock symbols (not brokerage holdings).

Returns the saved symbols, newest first, each with name, sector and market cap where available

  • add_to_watchlist — Use this when the user asks to add, watch or track Indian stocks in their DalalOS watchlist
  • remove_from_watchlist — Use this when the user asks to remove stocks from their DalalOS watchlist
  • submit_dalalos_feedback — Use this when the user asks to report a problem with DalalOS data or tools, or to request a feature, or agrees after you offer to report a likely DalalOS issue
  • search_stocks — Use this when the user names an Indian company, NSE symbol, BSE code or ISIN and you need to identify the listed security before calling another tool
  • get_stock_quote — Use this when the user asks for the price, trading volume, 52-week range, PE or PB of an Indian company listed on NSE or BSE
  • get_batch_quotes — Use this when the user wants end-of-day prices or ratios for several Indian stocks at once (up to 20), for example to line up peers.

Returns one item per symbol, each with its own status; the fields of get_stock_quote are under quote

  • get_financials — Use this when the user asks how an Indian company listed on NSE or BSE has performed financially: revenue, profit, margins, balance sheet, cash flow or growth, quarterly or annual
  • get_company_profile — Use this when the user asks what an Indian company listed on NSE or BSE is: its sector, industry, market cap and a headline financial summary.

Returns the company's identity and classification, its latest end-of-day market figures (price, market cap, P/E, P/B, 52-week range), a summary of the lates

  • get_company_report — Use this when the user asks a broad question about an Indian company listed on NSE or BSE ("how is X doing", "should I look at X"): a sourced report on quarterly and annual results, valuation, ownership, pledge, dividends and recent disclosures, no verdict
  • get_peer_comparison — Use this when the user compares an Indian company listed on NSE or BSE with similar companies on valuation, margins, return on capital or debt (PE, PB, ROE, debt-to-equity)
  • get_shareholding — Use this when the user asks who owns an Indian company listed on NSE or BSE: promoter, FII, DII, mutual fund and public holding percentages by quarter
  • get_price_history — Use this when the user wants past share prices of an Indian company listed on NSE or BSE over a period, daily, weekly or monthly, for example to chart or compare a move.

Returns end-of-day bars (open, high, low, close, volume) with 50- and 200-day averages

  • get_valuation_history — Use this when the user asks how an Indian stock's PE, PB (price-to-book, banks included), price-to-sales or market cap has changed over time.

Returns a series of P/E, P/B, price to sales, market cap and net debt, and for non-financial companies the enterprise-value multiples

  • get_returns — Use this when the user asks how much an Indian stock has returned over a period (1 week, 1, 3 or 5 years, or its full history), or about its volatility or drawdown.

Returns price returns for 1 week, 1, 3 and 6 months, 1, 3 and 5 years, the year to date and the full history, with annualised returns,

  • get_technical_indicators — Use this when the user asks for moving averages, RSI, MACD, Bollinger Bands, beta or circuit-limit data for an Indian stock.

Returns the latest values of the moving averages, RSI (14), MACD (12, 26, 9), Bollinger Bands (20, 2) and beta against the Nifty 50, with the stock's circuit-filter band and

  • get_delivery_trends — Use this when the user asks how much of an Indian stock's traded volume was taken for delivery, and how that has trended.

Returns, for each day, the shares traded, the shares taken for delivery and the delivery percentage, with the average over the window and over the last five sessions

  • get_dividends — Use this when the user asks whether and when an Indian company paid or declared a dividend, or about its dividend yield.

Returns each cash dividend with its ex-date, type and amount per share, the total for the last twelve months with the yield and payout ratio on it, and totals by financial year (

  • get_brsr — Use this when the user asks about an Indian company's BRSR (Business Responsibility and Sustainability Report) or ESG disclosures.

Returns the figures from the company's Business Responsibility and Sustainability Report as filed, by financial year, newest first: emissions, energy, water, waste, wor

  • get_bulk_block_deals — Use this when the user asks about bulk or block trades in an Indian stock on the NSE.

Returns the bulk and block deals NSE disclosed in the stock, newest first: date, client, buy or sell, quantity and price

  • get_insider_trades — Use this when the user asks which promoters, directors or company insiders bought, sold or pledged shares of Indian listed companies, market-wide or for one company
  • get_derivatives_positioning — Use this when the user asks about open interest, put-call ratio or futures and options positioning for an Indian stock or an index such as NIFTY or BANKNIFTY.

Returns the latest trading day's totals across all contracts of one F&O underlying: futures open interest and its change, call and put open

  • get_fo_participant_activity — Use this when the user asks how FIIs, DIIs, proprietary traders or clients are positioned in NSE futures and options across the whole market.

Returns one row per day and participant type (FII, DII, proprietary, client) with long and short contract counts in index and stock futures and options, acro

  • get_fno_ban_list — Use this when the user asks which Indian stocks are in the NSE futures-and-options ban period today.

Returns the securities in the F&O ban period on the latest day NSE published the list, with that date in ban_date

  • get_corporate_actions — Use this when the user asks about an Indian company's splits, bonuses, dividends, buybacks or other corporate actions, past or announced
  • get_stock_events — Use this when the user asks for the latest news on an Indian company, or what it has announced or filed recently: orders, board meetings, fundraising, acquisitions, management changes
  • get_glossary — Use this when the user asks what a term or ratio on an Indian company's page means or how it is calculated: P/E, ROE, TTM, gross NPA, promoter pledge and the like
  • get_company_documents — Use this when the user wants the documents an Indian company filed, grouped by type: results, annual reports, investor presentations, call transcripts and recordings, board meeting and AGM papers, ratings
  • search_company_disclosures — Use this when the user needs historical detail from an Indian company's earlier filings, for example a prior order value or commissioning date.

Returns passages from the company's filings that contain the search words, each with its page number, filing date and a link to the PDF

  • get_event_context — Use this after get_filing_extract when the user wants the earlier state behind an order, capacity or similar event in a filing.

Returns, for one filing given by news_id (from get_stock_events), the event recorded from it, the earlier event it follows, what changed between them, related disclosure

  • get_company_operating_metrics — Use this when the user asks for an Indian company's disclosed operating figures such as order book, order inflow, capacity or other business KPIs over time.

Returns the operating figures the company disclosed in its filings, each with its date, unit and source, and a coverage entry per metric say

  • get_company_macro_exposures — Use this when the user asks how an Indian company is exposed to a macro factor such as crude oil or the rupee-dollar rate, and whether the latest official move in it is large enough to matter.

Returns the company's reviewed exposures to macro factors that apply on as_of_date (ISO, default today),

  • get_company_exposures — Use this when the user asks which macro or commodity factors an Indian company depends on, based on reviewed exposure mappings.

Returns every reviewed exposure of the company to a macro or commodity factor: the factor, how it reaches the company (direct or indirect, direction, lag), any stated sens

  • get_relevant_peers — Use this when the user asks who an Indian company's real peers or competitors are and why
  • list_announcement_buckets — Use this when the user wants to browse market-wide disclosure types, such as order receipts or fundraises, before calling get_announcement_bucket.

Returns the kinds of disclosure that can be listed across the market: order-receipts, fundraise, credit-rating, acquisitions and management-changes

Notes

  • Description above is the publisher's own summary from their submission; check the website for plans, pricing and limits.
  • The endpoint answered a standard MCP initialize request when checked on 2026-11-13.

Sources: Website, Submission.

常见问题

1

DalalOS 远程 MCP 服务器是什么?

DalalOS 的远程 MCP 服务器是一个托管的 Model Context Protocol 端点,地址是 https://mcp.dalalos.in/mcp,AI 助手无需在本地安装或运行任何东西即可连接。

2

怎么连接 DalalOS 的 MCP 服务器?

把端点 https://mcp.dalalos.in/mcp 添加到任意支持 MCP 的客户端,比如 Claude Code、Cursor 或 VS Code。本页的接入代码块可以一步配置好每个客户端。

3

DalalOS 的 MCP 服务器需要认证吗?

不需要。DalalOS 的 MCP 服务器不需要认证——直接用端点地址连接即可。

4

DalalOS 的 MCP 服务器用的是哪种传输方式?

DalalOS 暴露的是 streamable-http 端点,这是远程 MCP 服务器普遍使用、主流客户端都支持的传输方式。

评论