get-content
ChatGPTFetch a single Inderes item by content ID (from list-content) or URL. Returns markdown — either the full body, or (for long reports) a table of contents you page through with read-document-sections. INPUTS (provide one): - contentId: Content ID from list-content (e.g. "COMPANY_REPORT:directus-9950") - url: URL from inderes.fi/se/dk (content page or direct PDF /files/... URL) OUTPUT (two shapes, told apart by documentId): - documentId is null → content is the item's full body as markdown (articles, comments, and reports served whole). - documentId is set → the item is long and split into sections: content is a short pointer, sections is the table of contents, and you read the parts you want by calling read-document-sections with that documentId and section numbers. Applies to research reports and ingested PDFs. EXAMPLES: Get article by ID: { contentId: "ANALYST_COMMENT:directus-1234" } Get content by URL: { url: "https://www.inderes.fi/fi/tutkimukset/raute-tutkimus-abc123" } Get PDF report directly: { url: "https://www.inderes.fi/files/c7cab272-d952-4e58-a47b-d13aa3feb15b" } Get content in specific language: { contentId: "COMPANY_REPORT:directus-9950", lang: "en" }
get-document
ChatGPTGet document metadata and table of contents by document ID. Use this tool when you have an ingested-PDF documentId from list-company-documents and need to see available sections before reading content. WORKFLOW: 1. search-companies → get companyId 2. list-company-documents with companyId → browse documents (paginated) 3. get-document with documentId → see sections/TOC 4. read-document-sections with documentId + sectionNumbers → read content Only ingested-PDF ids are accepted. A research report id from get-content (e.g. "COMPANY_REPORT:directus-10658#sv,en") does not work here and does not need to: get-content already returns that report's sections, so pass it straight to read-document-sections. EXAMPLE: { documentId: "cmkb7r6fc0000s9lk8jwzdth2" }
get-forum-posts
ChatGPTFetch posts from an Inderes forum thread. Provide threadUrl from search-companies.threadUrl or search-forum-topics.url. Pagination is Relay-style: use last to read the most recent posts (highest post numbers) and first to read the oldest posts (lowest post numbers). Exactly one of first or last is required. - Most recent posts: last: 50 (no cursor). - Oldest posts: first: 50 (no cursor). - Older page: { before: pageInfo.startCursor, last: N }. - Newer page: { after: pageInfo.endCursor, first: N }. Continue while hasNextPage / hasPreviousPage is true.
get-fundamentals
ChatGPTGet historical financial fundamentals for one or more companies. Returns income metrics, valuation multiples, margins, and profitability ratios. Supports field selection (use 'fields' to request only what you need) and year filtering. EXAMPLES: - Get all fundamentals for Nokia: { companyIds: ["COMPANY:200"] } - Get just P/E and dividend yield for 2022-2024: { companyIds: ["COMPANY:200"], fields: ["pe", "dividendYield"], startYear: 2022, endYear: 2024 } - Compare quarterly EBITDA margins: { companyIds: ["COMPANY:200", "COMPANY:201"], resolution: "quarterly", fields: ["ebitda", "ebitdaPercent"] }
get-inderes-estimates
ChatGPTGet forward-looking analyst estimates and recommendations for one or more companies. Each response includes recommendation (BUY/HOLD/SELL etc.), target price, and risk scores alongside the requested estimate fields. If companyIds is omitted, returns all covered companies (~150-200). yearCount controls how many future years are returned (near-term first); set includeQuarters=true to add quarterly breakdowns for those years. Common fields: revenue, revenueGrowth, ebitda, epsReported, dividend, dividendYield, pe, pb, sharePrice
get-model-portfolio-content
ChatGPTGet the current stock holdings in the Inderes model portfolio. Returns one row per stock position. acquisitionPriceEur is the total EUR amount originally used to buy the full position. currentValueEur is the current total EUR market value of all owned shares in that position. Use get-model-portfolio-price if you need the portfolio's historical total value over time instead of the current position breakdown.
get-model-portfolio-price
ChatGPTGet historical total portfolio value over time for the Inderes model portfolio. The priceHistory series contains the portfolio's total value for each date in EUR. The cash field is the current uninvested cash balance held in the portfolio, also in EUR. Defaults to monthly resolution over the full available history; narrow with dateFrom and/or switch to daily/weekly scale when a finer view is needed. Use get-model-portfolio-content if you need the current stock-level position breakdown instead of the historical total-value series.
get-transcript
ChatGPTGet a transcript by ID. Returns plain text with speaker labels — token-efficient format for LLM consumption. Use list-transcripts to find transcript IDs, then this tool to read the full text. EXAMPLES: { transcriptId: "VIDEO:directus-20205" } { transcriptId: "VIDEO:directus-20205", lang: "en" }
list-calendar-events
ChatGPTList company and market calendar events from Inderes. Use this for upcoming earnings dates, dividend events, general meetings, capital market days, and related events.
list-company-documents
ChatGPTList annual/interim reports and other filings issued by the company itself. Returns paginated documents with period info — use get-document to see sections/TOC for a specific document. For Inderes-authored reports, analyst comments, and articles, use list-content instead. Supports pagination via first (default 20, max 100) and after (cursor from pageInfo.endCursor). WORKFLOW: 1. search-companies → get companyId 2. list-company-documents with companyId → browse documents (paginated) 3. get-document with documentId → see sections/TOC 4. read-document-sections with documentId + sectionNumbers → read content EXAMPLE USAGE: {"companyId": "COMPANY:200"} {"companyId": "COMPANY:200", "first": 50, "after": "b2Zmc2V0OjIw"}
list-content
ChatGPTList content items from Inderes (reports, articles, analyst comments, releases); pass an item's id to get-content to read the body. Filter by company and/or content type. The returned id (e.g. "COMPANY_REPORT:directus-9950") is a content ID, not a document ID — it only works with get-content. To read PDF report sections, call get-content first; for ingested PDF reports it returns a separate documentId for read-document-sections. TIP: PRESS_RELEASE and STOCK_EXCHANGE_RELEASE are high-volume — filter them with the types parameter to avoid drowning out analyst research.
list-insider-transactions
ChatGPTList insider transactions from Inderes. Use this for insider buy/sell monitoring, governance signal tracking, and company-specific transaction history.
list-transcripts
ChatGPTList transcripts of earnings webcasts and analyst interviews. Returns metadata for browsing — use get-transcript to read the full text. EXAMPLES: Get latest transcripts: { first: 20 } Filter by company: { companyId: "COMPANY:200", first: 10 } Get next page: { first: 20, after: "cursor_from_previous_response" }
read-document-sections
ChatGPTRead the actual content of specific sections from a document by their section numbers. OUTPUT: - Returns the combined markdown content of all requested sections - Sections are retrieved based on the table of contents structure - Each section includes its header and all content until the next section at same or higher level A documentId reaches you one of two ways, and they differ in whether get-document is involved: - Ingested PDF (id looks like "cmkb7r6fc0000s9lk8jwzdth2"): search-companies → list-company-documents → get-document for the sections/TOC → read-document-sections. - Research report (id looks like "COMPANY_REPORT:directus-10658#sv,en"): get-content already returned both the documentId and its sections, so come straight here. Do not call get-document with this kind of id — it only accepts ingested-PDF ids. EXAMPLE USAGE: To read sections 1, 2, and 5 from a document: {"documentId": "cmkb7r6fc0000s9lk8jwzdth2", "sectionNumbers": [1, 2, 5]}
search-companies
ChatGPTSearch for companies by name (partial matches supported). Use this first to get a companyId — most other tools require one.
search-forum-topics
ChatGPTSearch Inderes forum topic titles and return up to 10 matching threads. Use this to find relevant thread URLs before calling get-forum-posts. This searches titles only, not post bodies. order controls ranking and defaults to relevancy.