Hunt for a job, skip the grind.

Find, track, and analyze opportunities from the terminal.

---

hunt is a Rust CLI that replaces the spreadsheet-and-bookmarks approach to job searching. It ingests job alerts from email, fetches full descriptions via browser automation, deduplicates postings, and uses AI to extract keywords, analyze fit against your resume, and generate tailored resume variants.

Features

• Email ingestion — Pull job alerts from Gmail (LinkedIn, Indeed) via IMAP
• Browser fetching — Scrape full job descriptions with Selenium/geckodriver, handle JS-rendered pages and "Show more" buttons
• Deduplication — Fuzzy title matching (Jaro-Winkler > 0.8), substring matching, URL dedup
• AI analysis — Analyze postings, extract domain-categorized keywords with weights, fit analysis, full resume tailoring
• Multi-model AI — Claude via CLI (subscription), direct Anthropic API, OpenAI API
• Resume management — Store base resumes, generate tailored variants per job/model/format
• Employer research — Track startup info, public company controversies, PE/VC ownership
• Ranking — Score jobs by pay, employer status, and application stage

Browse TUI

Quick start

# Build
cargo build --release

# Initialize the database
hunt init

# Add a job manually
hunt add "Senior DevOps Engineer at Acme Corp..."

# Or pull from email
hunt email --username you@gmail.com --password-file ~/.gmail.app_password.txt

# Fetch full descriptions from URLs
hunt fetch 5                    # single job
hunt fetch --all --limit 20     # batch

# List and rank
hunt list --status new
hunt rank --limit 10

AI-powered analysis

AI commands default to claude-sonnet via the Claude Code CLI (no API key needed if you have a subscription). Pass --model to switch.

# Analyze a job posting
hunt analyze 5

# Extract keywords (grouped by domain, weighted by importance)
hunt keywords 5
hunt keywords 5 --show           # view stored keywords without re-running AI
hunt keywords --search terraform  # search across all jobs

# Resume fit analysis
hunt fit 5 --resume devops-2026

# Generate a tailored resume
hunt resume tailor 5 --resume devops-2026
hunt resume tailor 5 --resume devops-2026 --models claude-sonnet,gpt-5.2
hunt resume compare 5            # compare variants side by side

Keyword output

Keywords are categorized into four domains with importance weights:

Keywords for job #5: Engineering Manager, Infrastructure Platform (model: claude-sonnet)

  *** = required   ** = important   * = nice-to-have

  TECH
    *** Kubernetes, Terraform
     ** PostgreSQL, Python
      * Airflow, Go, Snowflake, TypeScript, dbt

  DISCIPLINE
    *** DevOps, SRE, microservices
     ** CI/CD, Kanban

  CLOUD
    *** AWS

  SOFT SKILLS
    *** leadership, communication
     ** mentoring, planning

  PROFILE
  Tech-heavy infrastructure role. Strong emphasis on AWS/Terraform/Kubernetes
  stack with DevOps ownership. Leadership expectations high — managing a team,
  not just individual contribution.

Available models

Short name	Provider	Notes
claude-sonnet (default)	Claude CLI	Uses Claude Code subscription
claude-opus, claude-haiku	Claude CLI	Subscription billing
api-sonnet, api-opus, api-haiku	Anthropic API	Requires ANTHROPIC_API_KEY
gpt-5.2, gpt-4o, o3	OpenAI API	Requires OPENAI_API_KEY

Employer management

hunt employer list
hunt employer show "Acme Corp"
hunt employer block "Evil Inc"      # never apply
hunt employer yuck "Meh LLC"        # apply reluctantly
hunt employer ok "Redeemed Co"      # clear status

# Research
hunt employer research "Startup Co" # YC, funding, HN mentions
hunt employer evil "Big Corp"       # controversies, labor practices
hunt employer ownership "Acquired"  # parent company, PE/VC

Data management

# Cleanup
hunt cleanup --artifacts            # remove nav junk from email parsing
hunt cleanup --duplicates           # deduplicate jobs
hunt cleanup --all --dry-run        # preview all cleanup

# Destroy everything
hunt destroy                        # preview
hunt destroy --confirm              # actually wipe

Architecture

Module	Purpose
main.rs	CLI (clap), command dispatch, display formatting
db.rs	SQLite schema, migrations, all queries
ai.rs	AIProvider trait, three providers, prompt functions
models.rs	Data structs (Job, Employer, JobKeyword, etc.)
email.rs	IMAP ingestion, LinkedIn/Indeed HTML parsing
browser.rs	Selenium-based job description fetching

Database is stored at ~/.local/share/hunt/hunt.db (XDG). Schema auto-migrates on init or first use.

Building from source

Requires Rust 2024 edition. Warnings are treated as errors.

cargo build --release
ln -sf "$(pwd)/target/release/hunt" ~/.local/bin/hunt

For browser fetching, you also need geckodriver:

geckodriver --port 4444 &
hunt fetch 5