JobYap for AI Agents

JobYap's job postings, salaries, full descriptions and community discussions are available to AI agents through two read-only surfaces. No account, API key or authentication is required.

MCP server

A remote Model Context Protocol server (stateless streamable HTTP):

https://mcp.jobyap.com/mcp

Tools: search_jobs, get_job, get_job_comments, search_locations, list_companies, get_job_stats, plus search/fetch implementing the OpenAI deep-research contract.

• Claude Code: claude mcp add --transport http jobyap https://mcp.jobyap.com/mcp
• Claude.ai / Desktop: Settings → Connectors → Add custom connector → https://mcp.jobyap.com/mcp
• Cursor: Add to Cursor (one click), or add { "mcpServers": { "jobyap": { "url": "https://mcp.jobyap.com/mcp" } } }
• Windsurf / Zed / others: add { "mcpServers": { "jobyap": { "url": "https://mcp.jobyap.com/mcp" } } }
• VS Code: Install in VS Code (one click), or add { "servers": { "jobyap": { "type": "http", "url": "https://mcp.jobyap.com/mcp" } } }
• ChatGPT: Settings → Connectors (developer mode) → Add → https://mcp.jobyap.com/mcp, No Authentication
• Google Antigravity: MCP config with "serverUrl": "https://mcp.jobyap.com/mcp", or install the plugin from jobyap/agent-skills
• stdio-only clients: npx -y jobyap-mcp (bridges stdio to the hosted endpoint)

JSON API

The same capabilities as plain GET endpoints under https://jobyap.com/api/v1:

• GET /jobs: search by q (title keywords), company, location, work_mode, posted_within and sort, with cursor pagination
• GET /jobs/{id}: full posting with the description as markdown, salaries, locations and apply URL
• GET /jobs/{id}/comments: the discussion thread
• GET /locations?q=: resolve place names to filter identifiers
• GET /companies: every company on the JobYap Tech Index, with active-job counts and company-page slugs/URLs
• GET /stats: aggregate counts

Example: curl "https://jobyap.com/api/v1/jobs?q=engineer&work_mode=remote&posted_within=7d"

The API is intended for tools and agents rather than search-engine indexing, which is why robots.txt keeps /api/ non-indexable. Responses carry cache headers; please keep page sizes at or below 50 and avoid unbounded bulk requests, as the RSS feed and sitemap exist for bulk discovery.

Agent Skill

The jobyap skill teaches agents when and how to use this API, including the location-identifier grammar, pagination and search semantics:

npx skills add jobyap/agent-skills

Claude Code users can install the skill and the MCP server together as a plugin:

/plugin marketplace add jobyap/agent-skills
/plugin install jobyap@jobyap

Source, plugin manifests and every installer: github.com/jobyap/agent-skills. Machine-readable overview: llms.txt.