Skip to content

REFERENCE

Data sources and freshness

A market-data answer is only useful when you can tell where it came from and how old it is. DalalOS is designed around that constraint: market-data tools read a maintained cache populated from official and public Indian-market sources, and responses expose provenance and freshness instead of hiding cache age behind a single number.

This page is the canonical reference for the source and freshness semantics an AI client, developer or researcher should use when interpreting a DalalOS response.

KEY TAKEAWAYS

  • Stock prices and quote-style market data are end-of-day, not live ticks.
  • Company financials and disclosures are grounded in exchange/public filings rather than generated estimates.
  • Tool responses expose provenance and cache-age signals so a client can surface or gate stale data.
  • Read paths do not scrape upstream on demand; missing coverage stays explicit instead of being inferred.

Primary source families

Source familyWhat DalalOS uses it for
NSE archivesEnd-of-day prices, traded-market fields, index and exchange datasets where the NSE publishes an official archive or file.
BSE official interfacesSecurity-master and exchange-published company, filing, corporate-action and disclosure data used by the relevant DalalOS domains.
BSE-XBRL / SEBI-mandated filingsReported company financial statements and filing-derived line items used for financials and mechanically-computed ratios.
Official public-sector datasetsMacro and reference datasets exposed by DalalOS use the named official publisher for that domain, rather than a third-party market-data vendor.

Freshness is domain-specific

There is no honest single label such as “real-time” that applies to every market-data domain. End-of-day prices refresh after an exchange session is published. Company filings and corporate events arrive when the exchange publishes them. Ownership and other periodic disclosures update on their reporting cadence. A client should therefore use the freshness metadata returned with the requested domain rather than assume every field shares one timestamp.

How to read a DalalOS response

  • source identifies the response provenance or the DalalOS data contract that assembled it.
  • cached_at records when the relevant cached result was produced or refreshed.
  • stale tells the caller when cached data has exceeded the freshness policy for that domain.
  • coverage notes / reason codes distinguish missing, not-applicable or incomplete coverage from a valid zero.

When freshness itself is the question, use the get_data_freshness MCP tool. It reports the most recent successful refresh by data domain and the cache-health state the client should consider before relying on a figure.

What happens when upstream data is unavailable

The MCP read path does not fetch or scrape an exchange website during the user's tool call. Scheduled ingestion owns upstream collection; tools read the maintained cache. This means a transient upstream problem does not automatically turn a research question into an unbounded live scrape. Depending on the domain, the response can return cached data marked stale, an explicit no-data/not-found state, or a coverage note that says what is missing.

What DalalOS is not

  • Not a real-time tick or order-book feed.
  • Not a broker or order-routing service; market-data tools do not place trades.
  • Not an investment-advice or target-price engine.
  • Not a source of invented values when an exchange or filing does not disclose the requested fact.

Practical prompt pattern

Ask your AI client to include provenance whenever recency matters: “Use DalalOS to fetch the latest available quote and financial context for RELIANCE. Name the tools used, show the source, cached/freshness date and any stale or coverage warning. Do not substitute web search for a missing DalalOS value.”

FAQ

Common questions

Where does DalalOS get Indian stock-market data?

DalalOS uses official and public sources, including NSE archives for end-of-day market data, BSE official data interfaces and exchange filings, and SEBI-mandated BSE-XBRL filings for company financials. Some macro and reference-rate datasets come from their respective official Indian public-sector publishers.

Is DalalOS real-time?

No. DalalOS is a research data layer, not a live tick feed. Stock prices and quote-style market data are end-of-day, while filings, ownership, events and other domains update on their own scheduled or publication-driven cadence.

How can an AI assistant tell whether a DalalOS result is fresh?

DalalOS responses carry provenance and cache-freshness metadata. Clients should inspect the returned source, cached_at timestamp and stale status, and can call get_data_freshness when they need an overall view of cache health by data domain.

What happens when current data is unavailable?

DalalOS reads from its cache rather than scraping upstream during a tool call. A tool may return cached data marked stale, an explicit no-data/not-found state, or a coverage note instead of silently inventing a value.

Connect your AI to Indian stock market data

Sign in to DalalOS and connect your AI in one line of config. Free to start.

KEEP EXPLORING