MCP cookbook

These cookbook recipes show what you can ask Claude, Codex, Cursor, or another MCP client to do for your job board. You describe the outcome in ordinary language. The agent can inspect Cavuno’s live API schema, find the records it needs, make the change, and verify the saved result.

Start by connecting Cavuno to your AI client. You do not need to look up job IDs, post IDs, company IDs, or API paths before asking for help.

Manage jobs

• Post a job to your job board with AI — find or create the company, prepare the listing, and publish it when it is ready.
• Scrape and sync jobs from a careers page with AI — calibrate which roles belong, run the first import under supervision, then keep the source in sync.
• Build a custom job scraper for your job board with AI — turn a successful source-specific import into a repeatable Codex automation or scheduled workflow.

Grow your job board blog

• Research keywords for your job board with AI — find and group the searches that matter to your audience without forcing every keyword into a blog post.
• Find content gaps in your job board blog with AI — compare the complete existing Cavuno blog with audience and search demand to find missing or weak coverage.
• Plan your job board content strategy with AI — turn audience needs, business goals, keyword research, and content gaps into a prioritised publishing plan.
• Create and publish a job board blog post with AI — prove a new article should exist, define its information gain, draft it, and verify the published result.
• Improve internal linking in your job board blog with AI — find useful links between existing Cavuno posts, approve exact placements, and verify the updated articles.
• Refresh job board blog posts with Google Search Console — match Cavuno posts to first-party search data, select the right pages, and update only the approved posts.

Explore every supported resource

The recipes above are representative, not a separate feature set. The MCP reference covers the complete Operator API surface available through MCP: jobs, companies, blog posts, authors, tags, taxonomies, settings, domains, operations, audit logs, usage, and supported media metadata.

When a task requires a website, local file, spreadsheet, analytics platform, or third-party API, your AI client must supply that access separately. Cavuno MCP remains responsible for the Cavuno side of the task.