Job Application Tracker for developers
Everything you can do in Job Application Tracker, you can do from a script, a server or an AI agent — over a versioned REST API or over MCP.
The REST API answers at https://www.jobapplicationtracking.com/api/v1 and the MCP server at https://www.jobapplicationtracking.com/api/mcp. The machine-readable schema is at https://www.jobapplicationtracking.com/api/v1/openapi.json, and it needs no credential.
What the API covers
One registry of operations serves both surfaces, so anything offered over REST is offered as an MCP tool with the same input, the same output and the same permissions.
- Boards, applications and the status columns a card moves between.
- Companies, contacts, the interactions you log against them, and interviews.
- Insights, the spreadsheet table view, CSV import and export, and resumes.
- Your profile, and the boards you have shared with an organization.
- For coaches, colleges and workforce agencies: organizations, staff, cohorts, shared boards, the roster, audit events and billing.
Two environments
Every credential belongs to either live or sandbox. A live credential reaches your real job search; a sandbox credential reaches a seeded copy of the sample world that you can wreck and reset. Build against sandbox, then swap the key.
Two kinds of credential
- API keys
- For your own scripts and for MCP clients that let you set a static Authorization header. Created in Settings → Developers, scoped when you create them, and shown once.
- OAuth 2.1 access tokens
- For software acting on behalf of somebody else, and for MCP clients that discover authorization themselves — Claude.ai custom connectors among them. This app is its own authorization server.
Both travel as a bearer token in the Authorization header, and never in a query string. Authorization is exactly the web app’s: the API adds no second path to anybody else’s data.