Back to API Reference

Mira API

Update Log

User-facing version notes for the public Mira API contract. The top section summarizes the current release against the previous documented version; the history below records the supported public workflows by version.

Latest changes

Version 2.1.2 from 2.1.1

Current
  • Public job search and job detail workflows are now documented.
  • Public company search and company detail workflows are now documented.
  • Public search result windows are raised from 100 to 1000 results per request.
  • Entity detail endpoints accept up to 100 IDs or LinkedIn URLs per request.
  • Contact unlock is capped at 10 LinkedIn URLs per request with quota preflight before downstream unlock work.
  • Job data is documented as daily active postings; older job IDs can return not_found after postings close.
  • Profile and company references note quarterly 202603 data snapshots.
  • OpenClaw skills are aligned to the expanded public job and company workflows.
  • openjobs-platform-assistant is documented as the recommended public skill entrypoint for outcome-oriented recruiting workflows.
Version history

Documented API Versions

These entries describe the user-facing API contract for each documented public Mira API version.

2.1.2Current

Expands the public REST contract with job and company search/detail workflows, raises public search windows to 1000, expands entity detail batches to 100, and aligns the public OpenClaw skills with those capabilities.

  • POST /v1/job-fast-search returns string job_ids for active job postings.
  • POST /entity/v1/jobs/detail-by-id fetches full public job details from returned job_ids.
  • POST /v1/company-fast-search returns string company_ids for company records.
  • POST /entity/v1/companies/detail-by-id fetches full public company details from returned company_ids.
  • Public people, job, company, and scholar searches can request up to 1000 results, up from the previous 100-result window.
  • Entity detail endpoints accept up to 100 string IDs or LinkedIn URLs per request, up from the previous 50-item detail batch.
  • POST /v1/people-unlock accepts up to 10 LinkedIn URLs per request; quota is checked for the maximum possible request cost before downstream unlock work, then charged only for returned contacts.
  • Job data is documented as daily active postings, with not_found possible after a posting closes or leaves the active read model.
  • OpenClaw skills include openjobs-platform-assistant for workflows that combine jobs, companies, profiles, matching, and selected contact unlock.
2.1.1Previous

Introduces the ID-first entity detail model and public field-boundary documentation that 2.1.2 builds on.

  • Search workflows return stable string IDs before detail lookup.
  • Entity detail endpoints accept returned string IDs for downstream full-record retrieval.
  • Profile detail lookup supports profile IDs and LinkedIn URLs.
  • Detail examples show public _source defaults and supported public selections.
  • The unified response envelope remains { code, msg, data }.
  • Version discovery uses GET /version.
2.1.0Previous

Focused public API usage around stable string IDs, public detail fields, cleaner response formatting, and clearer version discovery.

  • Search results returned string profile and job IDs for downstream detail lookups.
  • Entity detail APIs accepted the string type IDs returned by search.
  • Public people and job search result windows were documented at 100 results per request.
  • Scholar search supported explicit result sizing up to 100 for public API keys.
  • Job detail defaults returned public job detail fields.
  • Source selection followed public field access rules for public API keys.
2.0.1Earlier

Published the ID-first public REST contract and matching OpenClaw skills for search, entity detail, grading, jobs, and scholars.

  • People and job search returned IDs first, with full documents fetched through entity detail APIs.
  • Profile detail lookup supported profile IDs and LinkedIn URLs.
  • Job detail lookup supported job IDs.
  • People grading, bulk grading, comparison, contact unlock, stats, job search, and scholar search were documented together.
  • Version discovery used /version.
Integration notes

Recommended Update Path

Use these notes when refreshing integrations from the previous public contract to the current one.

  • Add the new job and company search/detail endpoints if your integration needs active postings or company records.
  • Increase requested size values up to 1000 where broader public search windows are useful.
  • Raise entity detail client batch sizes up to 100 where larger detail hydration batches reduce round trips.
  • Keep unlock requests to 10 LinkedIn URLs or fewer and handle 402 preflight rejection when quota cannot cover the maximum possible unlock cost.
  • Handle not_found arrays from job detail because active job postings can close after search.
  • Install or refresh OpenClaw skills with npx skills install OpenJobsAI/openjobs-openclaw-skills when your client supports repository installs.
  • Use openjobs-platform-assistant as the default public skill entrypoint for workflow-style requests, then let it route to domain operations.
  • Keep parsing REST responses with the same { code, msg, data } envelope.