Skip to main content
Glama
669,041 tools. Updated 2026-10-17 09:22

"A server for unstructured data or information" matching MCP tools:

  • Creates a new submission for a specific Data Template (Form). Use this tool after obtaining the Data Template (Form) schema (via wdf_data_templates_get_schema_and_sample_submissions) and collecting all required information from the user. Keys in data must match the field IDs from the Data Template (Form) schema, not display labels — an unrecognized key may be silently dropped rather than raising an error. Provide all fields marked as required in the Data Template (Form) schema; omitted optional fields are simply left blank on the new submission.
    ConnectorNo auth
  • Get comprehensive account details for specific domain(s). USE FOR: Quick lookup of specific company information, engagement metrics, revenue data. DO NOT USE FOR: Bulk analysis or complex filtering (use get_insights_report instead). RETURNS: one entry per requested domain, in request order, each carrying its `domain`: engagement metrics (totalTouches, channels, websiteVisits), revenue data (openOppsAmount, closedWonOppsAmount), company info (name, industry, employees, engagementScore). A domain the platform could not answer comes back as {domain, error, data: []} in its own slot.
    ConnectorAPI key
  • Paid ($0.006 USDC). Pull the fields you define out of unstructured text as JSON. Anything the text does not state comes back null, never invented. Use to turn emails, invoices, contracts or listings into structured data. To get a web page's text first use genesis402_web_extract. The field descriptions steer the extraction, so include units and formats (e.g. "ISO date", "number in USD"). Returns one JSON object with exactly your keys, each holding the extracted value or null. Payment: call once without payment_signature to get the exact price quote (nothing is charged or signed), then call again with the signed payment to receive the result and settlement details.
    ConnectorNo auth
  • Get one court's identity, jurisdiction, profile URL, core metadata, and up to 10 judges classified as current when the user asks for court information or a roster. For a named judge's courtroom presentation or practice rules, use get_courtroom_presentation_rules. Use the court_url from find_court. Counts cover indexed profiles classified as active, senior, recalled, or visiting, not an independently verified complete bench. Retain roster coverage and location-specific contact notes. No pagination is available. An empty roster means no indexed current matches; a null count means unavailable roster data.
    ConnectorNo auth
  • Real-time web search via Tavily for current events, fact-checking, and research. Use search_depth='advanced' for complex queries (higher quality, higher cost) and topic='news' or 'finance' for headlines or market information. Use when: Choose when the task needs current, external, or factual information not available from on-chain or local data — e.g. news, prices, documentation, or fact-checking. Limitations: Returns web snippets, not raw page bodies; results depend on Tavily coverage. Advanced depth costs more. Not a substitute for on-chain tools like get_token_price. Alternatives: get_token_price, http_fetch
    ConnectorNo auth
  • Report that a public ContinueOps page or MCP tool response is incorrect, outdated, incomplete or confusing, so the team can fix it. Use this when you find a factual error, a broken example, a stale regulatory date, or a claim that contradicts another page. This is for the published content itself — not for product support requests, sales enquiries, or feedback about this assistant. Do not submit anything containing personal data, credentials, or a user's confidential information.
    ConnectorNo auth

Matching MCP Servers

Matching MCP Connectors

  • Turn documents into structured, AI-ready data by parsing, enriching, chunking, and embedding.

  • Oregon DMV MCP — live wait times at all 60 Oregon DMV field offices, plus the office

  • Get Reddit post by ID with its comments. Returns both the post data and comments in a single response. FAST (default, omit responseType or responseType="fast"): Returns post and up to 300 comments directly (use limit param to reduce, e.g. limit=5). Auto API fallback for fresh data. Results include guidance for full mode. PAGING (responseType="paging"): Async paginated results (100 comments/page), returns operationId for polling via checkOperationStatus. Supports pageNumber/tableName for subsequent pages. RESPONSE STRUCTURE: Returns { results: { post: {...}, comments: [...] }, count, guidance }. PAGING MODE DETAILS: FIRST CALL: Omit pageNumber and tableName. Creates cached table for comments, returns page 1 with post data and pagination metadata. SUBSEQUENT PAGES: Use tableName from first response with pageNumber (2, 3, etc.) to fetch additional comment pages. Post data is NOT returned on subsequent pages. FIELD SELECTION: Use postFields for post data optimization, commentFields for comment data optimization. First searches database for both post and comments, then external API if data is stale or missing. This is a safe, read-only tool for analyzing searchable information.
    ConnectorOAuth
  • Returns Fluentive's security, privacy, and compliance information. Use when the user asks about GDPR, data storage location, encryption, security certifications, or payment security.
    ConnectorNo auth
  • FREE. Service health and how recently the data was refreshed. Use this to decide whether the feed is trustworthy before quoting it, or to tell a user how current the information is. Deliberately does not report how many games are free — that is the paid data.
    ConnectorNo auth
  • Returns a complete pre-generated SAMPLE Phi Longevity PRISM report for a fixed SYNTHETIC persona (58-year-old male, type-2 diabetes). It is a fixed sample: it takes no input, analyzes nothing you send, and is not a real patient. For research/education with SYNTHETIC or de-identified data only. Do NOT submit protected health information (PHI). This endpoint is stateless and does not store inputs.
    ConnectorNo auth
  • Fetch detailed Google Maps information for a known `place_id`. Use google_map_search to find places or google_map_place_photos and google_map_place_reviews for the full photo and review data.
    ConnectorNo auth
  • Fetch detailed Google Maps information for a known `place_id`. Use google_map_search to find places or google_map_place_photos and google_map_place_reviews for the full photo and review data.
    ConnectorNo auth
  • Real-time web search via Tavily for current events, fact-checking, and research. Use search_depth='advanced' for complex queries (higher quality, higher cost) and topic='news' or 'finance' for headlines or market information. Use when: Choose when the task needs current, external, or factual information not available from on-chain or local data — e.g. news, prices, documentation, or fact-checking. Limitations: Returns web snippets, not raw page bodies; results depend on Tavily coverage. Advanced depth costs more. Not a substitute for on-chain tools like get_token_price. Alternatives: get_token_price, http_fetch
    ConnectorNo auth
  • Get Reddit subreddit by name with its posts. FAST (default, omit responseType or responseType="fast"): Returns subreddit data with up to 300 posts directly (use limit param to reduce). Auto API fallback for fresh data. PAGING (responseType="paging"): Async paginated results (100 posts per page), returns operationId for polling via checkOperationStatus. Supports pageNumber/tableName for subsequent pages. RESPONSE STRUCTURE: Returns { results: { subreddit: {...}, posts: [...] }, pagination: {...} }. FIELD SELECTION: Use subredditFields for subreddit data optimization, postFields for post data optimization. First searches database for both subreddit and posts, then external API if data is stale or missing. This is a safe, read-only tool for analyzing searchable information.
    ConnectorOAuth
  • Compare mapped human-evidence strength and limitations for two to five supplements for the same goal. This is an evidence comparison, not a product ranking or purchase recommendation, and contains no affiliate links. Use only for public, non-personal evidence questions. Do not call this tool for requests involving personal or sensitive health information, including medical records, medication lists, diagnoses, symptoms, laboratory results, or treatment planning. Tell the user not to submit that information and direct them to a qualified healthcare professional.
    ConnectorNo auth
  • Get detailed information for a specific VA facility. Returns comprehensive facility data including address, contact info, operating hours, services offered, patient wait times, and patient satisfaction scores. Args: facility_id: The VA facility ID (e.g., "vha_648", "vba_310", "nca_907"). Obtain from find_va_facilities results.
    ConnectorNo auth
  • Get detailed information about a nonprofit organization by EIN. Returns comprehensive data from the organization's IRS 990 filings including revenue, expenses, assets, executive compensation, and filing history. Use search_nonprofits first to find the EIN. Args: ein: Employer Identification Number (e.g. '13-1837418' or '131837418').
    ConnectorNo auth
  • Structured JSON extraction from unstructured text -- grounded in real, live verification, not just pattern-matching. Pulls requested fields, nulls anything missing, never guesses. Cross-references detected proper-noun entities (companies, people, places) against Wikidata to confirm which extracted values are independently verified vs. unconfirmed. SPENDS your balance -- requires authentication (OAuth). Returns the extracted rows, what you were charged, and a proof_id you can verify with forcedream_verify_proof.
    ConnectorNo auth
  • Intelligently match a free-text affiliation string (as found in academic paper bylines or grant records) to research organizations in the ROR registry. Pass the raw affiliation text exactly as it appears — e.g. "Dept. of Physics, Univ. of California, Berkeley, CA, USA" or "Max-Planck-Institut für Astronomie, Heidelberg". ROR's affiliation matching algorithm performs substring detection, name-variant lookup, and scoring to identify candidate organizations. Returns each candidate with: confidence score (0–1), a "chosen" boolean (the best single match), matching_type (e.g. SINGLE SEARCH, COMMON TERMS), and the organization's ROR ID, canonical name, type, and country. Ideal for metadata curation, ORCID affiliation disambiguation, funder acknowledgment normalization, and building organization-level research analytics from unstructured text.
    ConnectorNo auth
  • Free preview of a paid capability: checks whether useful data/analysis is available for a given input, before paying — proof of "I have information for this request," never the paid analysis itself. No payment, charge or account is ever involved. Not every capability supports preview (status is "unavailable" when it doesn't). Recommended flow: discover -> preview -> evaluate -> pay -> execute.
    ConnectorNo auth
  • Evaluates whether abstractions provide a genuinely different way of thinking or are structurally shallow. Use when adjacent layers feel redundant, wrappers add boilerplate without depth, or an abstraction feels leaky. Not for a single module's interface-to-implementation ratio (use deep-modules) or information leakage across boundaries (use information-hiding).
    ConnectorNo auth