sixtyfour_atlas_file
ChatGPTRead a bounded UTF-8 chunk from a path returned by Atlas inspect.
By SixtyFourAI
Deploy AI agents that investigate people and companies, resolve identities, map relationships, and surface risk signals — all through a single API.
struct for many companies at once — e.g. the companies you assembled from company searches. Provide the companies as either rows (a list of dicts of known identifiers like company_name, website, domain, linkedin) or csv_text (a CSV string with a header row), but not both. The server builds the upload file from your input. struct is required. Use find_people with people_focus_prompt and lead_struct for targeted people discovery, or full_org_chart for broad org visibility. The start response includes estimated_cost_credits, a floor estimate (extras like find_people can bill more). Poll the returned task_id with sixtyfour_enrichment_job_status; the completed response includes signed download URLs for the result files and the total charge_credits. This is a billable job — if a start call times out, do not blindly retry, since the job may already have been created. Practical note: companies passed inline travel in the model's context, so this tool is best for up to a few hundred rows assembled from searches. For larger user CSVs, use the Sixtyfour web dashboard or call the bulk API directly. Args: struct: Company output fields to research for every company. rows: Companies to enrich, one dict per company. Provide rows or csv_text, not both. csv_text: CSV text (with header) to enrich, as an alternative to rows. find_people: Whether to discover associated people for each company. people_focus_prompt: Criteria for people discovery when find_people is true. lead_struct: Output schema for each discovered person. full_org_chart: Return a compact org chart grouped by department. research_plan: Optional methodology for where/how to research. tier: Research depth/cost tier: micro, low, medium, or high. field_confidence: Return a confidence score per output field. columns: Subset of input columns to feed the agent. Defaults to all. result_formats: Saved result formats: "csv" (always produced) and/or "ndjson" (preserves typed values). save_json: Also save JSON output artifacts. webhook_url: Optional HTTPS webhook for completed results.target_company should contain identifiers such as company_name, website, domain, LinkedIn URL, or location. struct is required. Each key becomes a field in structured_data. Each value can be a plain-English description or an object like {"description": "Approximate employee count", "type": "int"}. Use find_people with people_focus_prompt for targeted people discovery. Use full_org_chart for broad department-level employee visibility. Example struct: { "company_name": "Name of the company", "industry": "Primary industry", "employee_count": {"description": "Approximate employee count", "type": "int"}, "uses_shopify": "True if the company uses Shopify" } Args: target_company: Company identifiers. struct: Company output fields to research and return. find_people: Whether to discover associated people. people_focus_prompt: Criteria for people discovery, e.g. "Find the CTO and VP of Marketing". lead_struct: Output schema for each discovered person. research_plan: Optional methodology for where/how to research. tier: Research depth/cost tier: micro, low, medium, or high. full_org_chart: Whether to return a compact org chart grouped by department.charge_credits, the amount charged for that page.result_handle (for get_company_details) and the response includes next_cursor and has_more (for next_page). Both expire after ~20 minutes. If a search returns 0 results, use sixtyfour_company_explore_field_values to check valid values for categorical fields. On 400 errors, call sixtyfour_company_get_searchable_fields to verify field names. To look up a company by domain, filter on website_domain. Linking to people: the company's string source_id and the people field currentCompanies.company.companyId represent the same LinkedIn company ID. Convert source_id to a JSON number for the people filter. Examples: {"industry": "SaaS"} {"employees_count": {"$gte": 100, "$lte": 5000}} {"hq_country_iso2": "US", "industry": "SaaS", "employees_count": {"$gte": 50}} {"industry": "SaaS", "$sort": {"employees_count": -1}, "$limit": 100} {"$or": [{"industry": "SaaS"}, {"industry": "Fintech"}]} {"ownership_status": {"$ne": "public"}} {"hq_country_iso2": {"$in": ["US", "GB", "DE"]}} {"description": {"$match": "cloud security"}} {"website_domain": "stripe.com"} {"funding_rounds": {"$elemMatch": {"name": "Series B", "amount_raised": {"$gte": 10000000}}}} {"linkedin_url": {"$exists": true}} Billable: 0.1 credit per result returned (a 100-row page costs 10 credits). Each response includes charge_credits, the amount charged for that page.task_id returned by any async enrichment start tool: sixtyfour_people_intelligence_start, sixtyfour_company_intelligence_start, sixtyfour_people_bulk_intelligence_start, sixtyfour_company_bulk_intelligence_start, sixtyfour_find_email_start, sixtyfour_find_phone_start, sixtyfour_reverse_email_start, or sixtyfour_reverse_phone_start. To cut the number of calls you make, pass wait_seconds (up to 25): the server blocks until the job is terminal or the wait elapses, then returns. While status is running, wait ~10s and call this tool again with the same task_id; only stop on a terminal status (completed, failed, cancelled, terminated, timed_out). Do not answer the user while the job is still running. Completed responses include the upstream result and charge_credits inside data. Bulk intelligence jobs return signed download URLs under data.results; non-bulk enrichment jobs return inline data under data.result and usually include task_type. Args: task_id: The task_id returned by an async enrichment start tool. wait_seconds: Seconds to block server-side until terminal (0 = return immediately, max 25).linkedin_urls (1-1000 profile URLs) or csv_text — the full raw text of the user's seed file. When the user points at a file or a vague list, read it and pass ALL of it rather than a few sampled URLs: more seeds produce a better similarity cohort. The cohort's defining traits (roles, skills, industry, seniority, ...) are inferred automatically; pass guidance ONLY for hard requirements the user explicitly stated. Async and billable: capture task_id, poll sixtyfour_find_similar_status until completed, then page ranked people with sixtyfour_find_similar_results. Seed profiles are always excluded from results.struct for many people at once — e.g. the leads you assembled from people searches. Provide the leads as either rows (a list of dicts of known identifiers like name, company, title, linkedin) or csv_text (a CSV string with a header row), but not both. The server builds the upload file from your input. struct is required. Each key becomes a field in the enriched output. Each value can be a plain-English description or an object like {"description": "Primary work email address", "type": "str"}. The start response includes estimated_cost_credits, a floor estimate. Poll the returned task_id with sixtyfour_enrichment_job_status; the completed response includes signed download URLs for the result files and the total charge_credits. This is a billable job — if a start call times out, do not blindly retry, since the job may already have been created. Practical note: leads passed inline travel in the model's context, so this tool is best for up to a few hundred rows assembled from searches. For larger user CSVs, use the Sixtyfour web dashboard or call the bulk API directly. Args: struct: Output fields to research for every lead. rows: Leads to enrich, one dict per lead. Provide rows or csv_text, not both. csv_text: CSV text (with header) to enrich, as an alternative to rows. research_plan: Optional methodology for where/how to research. tier: Research depth/cost tier: micro, low, medium, or high. field_confidence: Return a confidence score per output field. columns: Subset of input columns to feed the agent. Defaults to all. result_formats: Saved result formats: "csv" (always produced) and/or "ndjson" (preserves typed values). save_json: Also save JSON output artifacts. webhook_url: Optional HTTPS webhook for completed results.result_handle from a search result item. Expires after ~20 minutes; rerun search if expired. include_fields: Field names from sixtyfour_people_get_searchable_fields.lead_info should contain any known identifiers such as name, email, company, title, location, website, or LinkedIn URL. struct is required. Each key becomes a field in structured_data. Each value can be a plain-English description or an object like {"description": "The individual's primary work email", "type": "str"}. Example struct: { "name": "The individual's full name", "email": {"description": "Primary work email address", "type": "str"}, "linkedin": "LinkedIn URL for the person", "github_notes": "Detailed notes on their GitHub profile" } Args: lead_info: Lead identifiers. struct: Output fields to research and return. research_plan: Optional methodology for where/how to research. tier: Research depth/cost tier: micro, low, medium, or high.next_cursor value from a sixtyfour_people_search or sixtyfour_people_next_page response. Billable: 0.1 credit per result returned (a 100-row page costs 10 credits). Each response includes charge_credits, the amount charged for that page.result_handle (for get_person_details) and the response includes next_cursor, has_more, and total_available when the total is exact. Capped totals use total_available_lower_bound and appear as N+ in the summary. Set count_only=true to return only the exact match count and timing. Result handles and cursors expire after ~20 minutes. Args: query: MongoDB-style filter dict. Bare values imply $eq. Operators: $ne, $gt, $gte, $lt, $lte, $in, $nin, $exists, $match, $phrase, $elemMatch, $or, $and, $not, $nor, $search, $sort, $limit. page_size: Results per page. max_results: Cap on total results across all pages. include_fields: Field names to hint the preview planner. count_only: Return only the exact match count and timing. Ignores page_size, max_results, include_fields, $limit, and $sort. If a search returns 0 results, use sixtyfour_people_explore_field_values to check valid values. On 400 errors, call sixtyfour_people_get_searchable_fields to verify field names. Linking to companies: the long field currentCompanies.company.companyId is the numeric form of the company's string source_id. Use $elemMatch to filter people by company, and inside it use the full dotted field path with a JSON number. Use exact field names from sixtyfour_people_get_searchable_fields; the examples below use common ones (job title lives at currentCompanies.positions.title, not a top-level field). Examples: {"headline": {"$match": "engineer"}} {"currentCompanies.positions.title": {"$match": "engineering manager"}} {"currentCompanies": {"$elemMatch": {"currentCompanies.company.companyId": 1234567}}} Billable: 0.1 credit per result returned (a 100-row page costs 10 credits). Each response includes charge_credits, the amount charged for that page.email field. Poll the returned task_id with sixtyfour_enrichment_job_status. Args: leads: Leads to process (1-100), each with an email field.phone field. Poll the returned task_id with sixtyfour_enrichment_job_status. Args: leads: Leads to process (1-100), each with a phone field.subject should contain any known identifiers such as name, company, LinkedIn URL, email, or location — more is better.wait_seconds (up to 25): the server blocks until the job is terminal or the wait elapses, then returns. While status is running, wait ~10s and call this tool again with the same task_id; only stop on a terminal status (completed, failed, cancelled, terminated, timed_out). Do not answer the user while the job is still running. Completed responses include date_of_birth (day/month/year/confidence), age (value/as_of/confidence), evidence, and sources directly under data — this is a dedicated envelope, not the generic enrichment job-status shape.Click any prompt to copy it.
Add this badge to your site or README. It links to this page.
HTML
<a href="https://mcp-app-store.keenethics-labs.com/app/plugin_asdk_app_6a514c90130c8191bb9a04df335d5ac6"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>
Markdown
[](https://mcp-app-store.keenethics-labs.com/app/plugin_asdk_app_6a514c90130c8191bb9a04df335d5ac6)