recruitlayer

Use this when the user wants one job's complete record by job_id. Do not use this with a submission id; use get_pending_job_submission for receipts. Returns the job's details, postings, posting-product identity, click_count and application_count, apply-email recipients, and unresolved submission receipts. Works for every available job regardless of how it was created; unknown ids fail with not found.

get_job

Use this when the user wants to check whether an asynchronous job submission has posted. Do not use this to read the job itself; once status is 'posted', pass the receipt's job_id to get_job. Returns id, status (queued|processing|posted|failed), company_id, job_id and job_posting_id (null until posted), last_error, and created_at. The input is the submission id from create_job, not a job_id; unknown ids fail with not found.

get_pending_job_submission

Use this when the user needs the current distribution-product configuration for one company or market. With company context, each item includes the number currently available; with market context, it returns configuration metadata only. Exactly one context ID is required. It does not return monetary data or change state.

get_product_catalog

Use this when the user wants one sales lead row by sales_lead_id. Do not use this with a company_id; use get_company for company records. Returns the same fields as a list_sales_leads row (lead_state, origin, hot flag, typed inbound inquiry context, score, active signal types, and 24h/7d job counts) — no additional detail beyond the list item. Out-of-territory leads fail with not found.

get_sales_lead

Use this when the user wants the current measured counts and account-metric coverage for their sales territory. Do not use this to list the records behind a count; use search_company_account_ownerships, list_sales_leads, search_jobs, list_job_posting_product_expirations, list_job_postings_with_zero_applications, or get_candidate_pipeline_summary for those records or focused totals. Returns user_id, client_id, corporation_id, market_ids, employer_account_count, employer_account_metrics_count, employer_account_missing_metrics_count, lead_count, open_job_count, near_renewal_account_count, job_posting_product_expiration_count_48h, zero_application_job_posting_count, open_candidate_count, high_match_count, recent_application_count, metrics_source, and refreshed_at. Every count is measured over the accounts represented by employer_account_metrics_count. When employer_account_missing_metrics_count is above zero, report the coverage because the counts under-report the full territory.

get_sales_territory_summary

Use this when the user wants currently active campaign schedules for owned accounts. Do not use this for all campaign records; use list_campaigns instead. Returns {job_campaigns, has_more, next_offset}; each row carries campaign identity, company, start/end schedule, and overview link. company_id optionally narrows to one owned account and out-of-scope ids fail with not found.

list_job_campaign_statuses

Use this when the user wants to page through campaign schedules. Do not use this for active campaign schedules only; use list_job_campaign_statuses instead. Returns {campaigns, has_more, next_offset} newest-first; each row carries identity, company, schedule, and an overview link. Page with limit and offset.

list_campaigns

Use this when the user wants to page through available companies in creation order. Do not use this to find a company by name or location; use search_companies instead. Returns {companies, has_more, next_offset} ordered by creation time. Page with limit and offset; an empty page means no further companies.

list_companies

Use this when the user wants accounts with posting products that expire within 48 hours. Do not use this for the full account list or ownership; use search_company_account_ownerships instead. Returns {owned_company_accounts, has_more, next_offset} ordered by soonest product expiry; each row carries company identity, expiring_job_posting_count_48h, evaluated_at, and an overview link. An empty page means no covered account has a qualifying expiry.

list_company_accounts_needing_attention

Use this when the user wants which owned-account jobs have posting products expiring within 48 hours. Do not use this for an account's overall attention state; use list_company_accounts_needing_attention instead. Returns {job_posting_product_expirations, has_more, next_offset}; rows carry job_posting_product_id, job_posting_id, job_id, company_id, title, expires_at, and application_count. company_id optionally narrows to one owned account and out-of-scope ids fail with not found; pass a row's job_id to get_job.

list_job_posting_product_expirations

Use this when the user wants live owned-account jobs that have received no applications. Do not use this for candidate counts; use get_candidate_pipeline_summary instead. Returns {job_postings, has_more, next_offset}; rows carry job_posting_id, job_id, company_id, title, and application_count (always 0 here). company_id optionally narrows to one owned account and out-of-scope ids fail with not found; pass a row's job_id to get_job.

list_job_postings_with_zero_applications

Use this when the user wants the receipts of jobs submitted through this API. Do not use this to list actual jobs — jobs created in the hiring portal or by feeds never appear here; use search_jobs instead. Returns {pending_job_submissions, has_more, next_offset} newest-first; each receipt carries id, status, company_id, and job_id once posted. Page with limit and offset.

list_pending_job_submissions

Use this when the user wants the markets available for company, product, and search operations, or needs a market_id for another tool. Do not use this for one market's product configuration; use get_product_catalog instead. Returns {markets, has_more, next_offset} sorted by display name; each row carries market_id, name, display_name, corporation_id, and is_active. Page with limit and offset; an empty page means no further markets.

list_markets

Use this when the user wants sales leads in the territory, including un-onboarded prospects. Do not use this for accounts already owned; use search_company_account_ownerships instead. Returns {sales_leads, has_more, next_offset} ordered by lead priority: hot inbound employer inquiries newest-first, then score and recency. A lead's company_id is null until the employer is onboarded, lead_state is 'prospect' or 'claimed', and inbound rows include typed contact, hiring-needs, source, and market context. cbsa_code narrows to a metro area — a geographic area, not a platform market.

list_sales_leads

Use this when the user wants to find job applications — candidates in hiring pipelines — by text or filters. A candidate is an application to a company's job, not a global jobseeker profile. Do not use this for aggregate pipeline counts; use get_candidate_pipeline_summary instead. Returns {items, limit, offset, has_more, next_offset, match_count, is_truncated}; items carry name, stage, source, job/company ids and dates but never contact details — get_candidate returns those. is_truncated=true means more ranked matches exist beyond the bounded window; narrow the query or filters rather than paging further.

search_candidates

Use this when the user wants to find a company by name, market, corporation, description, or location — including resolving a company name to the company_id other tools take. Do not use this to read one known company_id; use get_company instead. Returns {items, total, limit, offset, has_more, next_offset}; total is the full match count. With q results rank by relevance; without q this browses newest-first with the filters applied. Results omit inactive companies, so is_active=false returns no matches; q needs at least 2 characters.

search_companies

Use this when the user wants to know who owns a company account, which accounts a person owns, or which accounts are unassigned. Do not use this to find a company or person by name; resolve names to ids with search_companies and search_users first, and use list_company_accounts_needing_attention for renewal risk. Returns {company_account_ownerships, has_more, next_offset}; each row carries the company identity/contact fields, assignment_status ('assigned' or 'unassigned'), and nullable sales_rep and sales_manager entries of user_id, name, and email. company_id filters to one exact company, owner_user_id matches either the assigned sales rep or the sales manager, and combining owner_user_id with assignment_status='unassigned' is rejected.

search_company_account_ownerships

Use this when the user wants to list, find, or count actual jobs — including how many open jobs a company has: pass company_id and status='open' and read total. Do not use this for API submission receipts; use list_pending_job_submissions instead. Returns {items, total, limit, offset, has_more, next_offset}; items carry title, status, company, location, salary, and published_at. With q results rank by relevance; without q, newest-published first. All filters combine with AND; jobs without salary data are excluded when a salary bound is set, and q needs at least 2 characters.

search_jobs

Use this when the user asks how recruitlayer itself works — features, workflows, or product questions answerable from documentation. Do not use this for live business data such as companies, jobs, candidates, campaigns, or accounts; use the matching search or get tool instead. Returns {question, status, passages}; each passage carries source_path, content, and resolved links. Answer only from the returned passages, cite each passage's source_path, and include relevant passage links as clickable Markdown links using their provided labels and URLs; when status is 'no_match', say the available documentation does not contain an answer. A retrieval failure is a tool error, not evidence of a documentation gap.

search_recruitlayer_documentation

Use this when the user wants to find a platform user — including resolving a person's name or email to the user_id other tools take, such as search_company_account_ownerships. Do not use this for people applying to jobs; use search_candidates instead. Returns {items, total, limit, offset, has_more, next_offset}; total is null because the directory does not run a count query. Each item carries user_id, first_name, last_name, and email. q matches first name, last name, and email with a 2-character minimum; omit q and role to browse the available directory.

search_users

Use this when the user wants to post a new job. Do not use this to edit an existing job; use update_job instead. Street Address and location are optional for every workplace mode. If supplied, location_city, location_state, and location_zipcode must be provided together. DC is a valid location_state. Returns a submission receipt, not the job: id (the submission id), status (queued at first), company_id, and null job_id/job_posting_id until posting completes. Posting is asynchronous — poll get_pending_job_submission with the receipt id, then pass its job_id to get_job. Retrying with the same idempotency_key and payload returns the original receipt.

create_job

Use this when the user wants to change a company's profile fields: display_name, contact and address details, url, description, external_id, or social ids. Do not use this to take a company offline or restore it; use deactivate_company or activate_company instead. Returns the updated company record; only fields present in the payload change and omitted fields keep their current values. Unavailable company_id values fail with not found.

update_company

Use this when the user wants to change editable fields on a saved job. Do not use this to post a new job; use create_job instead. Returns the refreshed complete job record; only supplied fields change. A supplied notification_recipients list replaces the whole recipient set, so switching to apply_url or apply_text requires sending an empty notification_recipients list in the same request. Set confirm_live_edit=true when materially changing a job with an open posting, or the call is refused; unavailable ids fail with not found.

update_job