Step 1 — Research prompt generator

Generates one ready-to-copy research prompt per target country, scoped to your role and profile. Claude does not estimate salaries itself — every number must come from research you supply.

Flow 

flowchart TD
  Start([Step 1 begins]) --> Handoff{Single country\nhanded off from\nCountry Finder?}
  Handoff -->|yes| UseCountry[Use that country\nSkip country list question]
  Handoff -->|no| AskList[Ask for target\ncountry list]
  UseCountry --> ReadProfile[Read profile.md\nfor role and seniority]
  AskList --> ReadProfile
  ReadProfile --> ForEach[For each country:\ngenerate research prompt]
  ForEach --> Prompt[Prompt instructs researcher to find\nlocal-market salary only\nexcluding expat · FAANG · US-skewed\ncontractor · equity-heavy data]
  Prompt --> Tiers[Request two company tiers:\nMid-size / Mainstream Local-Market\nPremium / International / Remote-first]
  Tiers --> More{More\ncountries?}
  More -->|yes| ForEach
  More -->|no| Done([Output ready-to-copy prompts\nWait for research results])

What it reads

  • profile.md — used to fill in your target role, seniority, and skills in each prompt
  • target_country from the state file — if set by a Country Finder handoff, the country list question is skipped

Country list

If invoked standalone (not handed off from Country Finder), Claude asks for your target country list before generating any prompts.

If invoked with a single country handed off from Country Finder, that country is used directly and the question is skipped.

Prompt content

Each generated prompt instructs the researcher to:

  • Use current web research only, prioritising sources from the last 12 months
  • Find realistic local-market annual base salary for your role and seniority
  • Target local candidates with a similar profile — not expat, FAANG-only, or global remote rates
  • Separate results into two company tiers:
    • Mid-size / Mainstream Local-Market — typical local employers
    • Premium / International / Remote-first — higher-paying segment
  • Provide a national range and major tech hub range if salary varies significantly by city
  • Include practical Low, Realistic, and Strong figures within each tier
  • Cite sources with dates

Excluded from research:

  • levels.fyi and FAANG-only data
  • Glassdoor US or US-skewed compensation
  • Inflated global or remote compensation
  • Contractor or freelance rates
  • Equity-heavy total compensation

Output

One ready-to-copy research prompt per country, clearly separated by country name. Claude waits for you to run the research and paste results back in Step 2.

Claude Compass — MIT License