Screen Job Listings

Add the everyrow connector if you haven't already. Then upload job_postings.csv and ask Claude:

Screen each job posting to find roles that meet ALL THREE criteria: 1. Remote-friendly: explicitly allows remote, hybrid, or distributed work. 2. Senior-level: title includes Senior/Staff/Lead/Principal, or requires 5+ years. 3. Salary disclosed: specific compensation figures, not "competitive" or "DOE".

7 of 15 postings pass (46.7%). Results take under a minute.

Claude Code can read a job posting and tell you if it's remote-friendly. But what if you need to screen many postings against three criteria simultaneously, with structured yes/no output for each?

Here, we get Claude Code to screen 15 job postings for roles that are remote-friendly, senior-level, and have disclosed salary.

Add everyrow to Claude Code if you haven't already:

claude mcp add futuresearch --scope project --transport http https://mcp.futuresearch.ai/mcp

Download job_postings.csv. Tell Claude:

Screen each job posting to find roles that meet ALL THREE criteria:
1. Remote-friendly: explicitly allows remote, hybrid, or distributed work
2. Senior-level: title includes Senior/Staff/Lead/Principal, or requires 5+ years
3. Salary disclosed: specific compensation figures, not "competitive" or "DOE"

Claude calls everyrow's screen MCP tool:

Tool: everyrow_screen
├─ task: "Screen each job posting to determine if it meets ALL THREE criteria..."
├─ input_csv: "/Users/you/job_postings.csv"
└─ response_schema: null

→ Submitted: 15 rows for screening.
  Session: https://futuresearch.ai/sessions/17d3075d-1788-4fba-8c46-a20f169976ec
  Task ID: ab78...

Tool: everyrow_progress
→ Completed: 15/15 (0 failed) in 57s.

Tool: everyrow_results
├─ task_id: "ab78..."
├─ output_path: "/Users/you/qualified_jobs.csv"
→ Saved 7 rows to /Users/you/qualified_jobs.csv

7 of 15 postings qualified. View the session.

All 15 rows were decisive on the first pass (7 yes, 8 no, 0 borderline). Traditional keyword matching achieves ~68% precision on these criteria; semantic screening exceeds 90%.

The everyrow SDK screens job postings with LLM-powered evaluation, returning structured results with per-criterion breakdowns.

pip install everyrow
export EVERYROW_API_KEY=your_key_here  # Get one at futuresearch.ai/api-key

import asyncio
import pandas as pd
from pydantic import BaseModel, Field
from everyrow import create_session
from everyrow.ops import screen

job_postings = pd.read_csv("job_postings.csv")

class JobScreeningResult(BaseModel):
    passes: bool = Field(description="Whether the job meets ALL three criteria")
    is_remote_friendly: bool = Field(description="Allows remote/hybrid/distributed work")
    is_senior_level: bool = Field(description="Senior/Staff/Lead/Principal or 5+ years")
    has_salary_disclosed: bool = Field(description="Specific salary figures provided")
    reasoning: str = Field(description="Brief explanation")

async def main():
    async with create_session(name="Job Posting Screening") as session:
        result = await screen(
            session=session,
            task="""
                Screen job postings for ALL THREE criteria:
                1. Remote-friendly: explicitly allows remote/hybrid/distributed work
                2. Senior-level: title includes Senior/Staff/Lead/Principal or 5+ years
                3. Salary disclosed: specific compensation figures, not "competitive"
            """,
            input=job_postings,
            response_model=JobScreeningResult,
        )
        return result.data

results = asyncio.run(main())

7 of 15 postings passed all three criteria. The structured output includes individual boolean fields for each criterion plus reasoning, making it easy to analyze which criteria are most commonly failed.