Advanced: Claude Code

• Claude Code can have unintended consequences if you can’t understand the code it writes, so I’m keeping the session focused on read-only reporting use cases.
• If you want to follow along, install the Claude Code desktop app beforehand.
• This is an advanced workflow: I’d only suggest joining if you’ve already run into limits with the MCP server. For most customers, the MCP connector is the easier and more secure way to automate with Stardex, and we always recommend starting there.

​

What you’ll get

• A spreadsheet of all actual compensation records across your database (name, title, job, total comp)
• Grouped averages by job title with a visual chart
• (Optional) AI-scored segments per candidate: seniority level, market tier, and profile grade

​

Step 1 — Download the starter folder

1. Download and unzip the Stardex + Claude Code starter folder.

2. Open the .env file inside it in any text editor.

Can’t see .env on Mac? In Finder, press Cmd + Shift + . (period) to show hidden files.

3. Get your Stardex API key:
   • Go to Settings → Email & other integrations → API access & Webhooks in Stardex (direct link: stardex.com/settings?tab=email-calendar&section=api-access)
   • Create a new key and copy it
   • Paste it into .env replacing paste_your_key_here:

STARDEX_API_KEY=paste_your_key_here

Only admin users can create API keys. If you don’t see API access & Webhooks, ask your Stardex admin.

​

Step 2 — Open Claude Code

1. Open Claude Code (desktop app or claude in terminal).
2. Click Open folder and select the starter folder you just unzipped.

​

Step 3 — Pull your compensation data

“Pull all actual compensation data across my database. For each record include the person’s name, current job title, job name if it’s linked to a job posting, and total comp.”

• Page through your entire database
• Filter to compensation_type = actual only
• Save a compensation_data.tsv file you can drag straight into Google Sheets

​

Step 4 — Segment by job title

“Group all [VP of Sales] titles together — normalize the variants — and show me the average, min, and max comp. Make a chart.”

• Collapse title variants like “Vice President of Sales”, “VP, Sales”, “VP of Sales - West” into one group
• Return a table with count, average, min, and max
• Generate a distribution chart saved as a .png

​

Step 5 — AI profile evaluation (optional)

STARDEX_API_KEY=your_stardex_key_here
ANTHROPIC_API_KEY=your_anthropic_key_here

“Evaluate each [VP of Sales] profile and segment them by seniority, location, and current industry. Export to a CSV.”

​

Example output

Name               Company          Comp      Seniority      Tier    Industry
Candidate A        Example Co. A    $200,000  Rising VP      Tier 1  SaaS
Candidate B        Example Co. B    $250,000  Rising VP      Tier 1  FinTech / Payments
Candidate C        Example Co. C    $300,000  Rising VP      Tier 1  Data / Analytics

​

Tips

• Re-run anytime — just type the same prompts again to get fresh data as your database grows
• Change the title group — swap in any title at step 4/5 to analyze a different role
• Read-only by default — Claude will never modify your Stardex data unless you explicitly confirm it