applyr

CLI job application tracker designed for AI coding agents.

Track applications, measure your conversion funnel, spot skill gaps, and generate ATS-optimized CVs — all from your terminal. Built to work with Claude Code, Cursor, Aider, or any AI coding agent.

Why applyr?

I built this while applying to 200+ jobs. Most job trackers are web apps that don't talk to your AI tools. applyr is different:

• CLI-first — runs in your terminal, pipes into anything
• AI-agent native — your coding agent analyzes offers, scores compatibility, and generates CVs
• ATS-safe CVs — locked CSS template that passes Applicant Tracking Systems. Your agent fills content, never touches the structure
• Zero dependencies — Python 3.12+ stdlib only. No frameworks, no API keys, no subscriptions
• Local and private — your data stays in a SQLite file on your machine

How it works

1. You paste a job offer into your AI agent
2. The agent reads your cv-master.md (your complete professional profile)
3. The agent evaluates your compatibility per topic (tech stack, experience, etc.)
4. The agent runs `applyr add` to register the offer with all data
5. If you want to apply, the agent generates an ATS-safe CV tailored to the offer
6. You track everything: pipeline, stats, follow-ups, skill gaps, trends

applyr is the storage and structure layer. Your AI agent is the brain that analyzes and decides.

---

Install

pip install applyr

Or clone and install locally:

git clone https://github.com/DeibyGS/applyr.git
cd applyr
pip install .

---

Setup (3 steps)

Step 1 — Initialize

applyr init

This creates ~/.applyr/ with:

• applyr.toml — configuration (scoring weights, thresholds, paths)
• jobs.db — SQLite database (empty, ready to use)
• cv-master.md — template for your professional profile
• AGENT_INSTRUCTIONS.md — step-by-step guide for your AI agent
• cv/ — directory for generated CVs

Step 2 — Fill your CV master

Open ~/.applyr/cv-master.md and fill it with your complete professional profile: contact info, experience, projects, skills, education, certifications, languages. This is the source of truth — the AI agent reads this file to evaluate offers and generate CVs. Never leave it empty.

Step 3 — Connect your AI agent

Copy the agent instructions into your AI tool's config:

# Claude Code
cat ~/.applyr/AGENT_INSTRUCTIONS.md >> ~/.claude/CLAUDE.md

# Cursor
cat ~/.applyr/AGENT_INSTRUCTIONS.md >> .cursorrules

# Aider / OpenCode / others
# Add the content to whatever file your agent reads for instructions

This tells your AI agent:

• How to read your cv-master.md and evaluate offers
• How to build the JSON for applyr add with all valid field values
• Which command to run for each user question
• Rules: never invent content, be honest with scores, leave unknown fields empty
• ATS rules for CV generation (single column, standard fonts, visible URLs, etc.)

That's it. You're ready.

---

Usage

Register an offer

Paste a job posting into your AI agent and say "analyze this offer". The agent will:

1. Read your cv-master.md
2. Score each topic (tech stack, experience, education, etc.)
3. Run applyr add with all the data:

applyr add '{"title": "AI Engineer", "company": "Acme Corp", "work_mode": "remote", "location": "Madrid", "salary_min": 30000, "salary_max": 40000, "seniority_level": "junior", "role_category": "ai", "tech_stack": "Python, LangChain, AWS", "canal": "linkedin_easy", "status": "applied", "topics": {"tech_stack": {"score": 85, "detail": "Python strong"}, "education": {"score": 70, "detail": "DAM completed"}, "english": {"score": 60, "detail": "B1"}, "experience": {"score": 40, "detail": "6mo internship"}, "projects": {"score": 90, "detail": "3 production projects"}, "cultural_fit": {"score": 80, "detail": "Good fit"}}}'

Output:

Offer added successfully.
  ID          : 1
  Title       : AI Engineer
  Company     : Acme Corp
  Compat.     : 74%
  Status      : Applied
  Follow-up   : 2026-08-16
  Skill gaps  : English, Experience

All fields are optional except title. The agent fills what it can from the posting.

View your offers

applyr list                    # All offers (last 50)
applyr list --status applied   # Filter by status
applyr show 1                  # Full detail of offer #1
applyr pipeline                # Grouped by status

Track your progress

applyr stats                   # Conversion funnel + metrics
applyr gaps                    # Skills you need to improve
applyr followups               # Overdue and upcoming follow-ups
applyr trends                  # Applications per week + growth rate
applyr summary --json          # Weekly summary as structured JSON

Update and manage

applyr update 1 waiting --notes "Interview scheduled for Monday"
applyr update 1 rejected --notes "They needed 3+ years experience"
applyr search Python           # Search by company/title/tech/notes
applyr delete 5                # Remove an offer
applyr export --format json    # Export everything

Generate ATS-safe CVs

applyr cv generate 1           # Creates HTML skeleton for offer #1

This generates an HTML file with:

• Locked ATS-safe CSS — single column, standard fonts, no flex/grid/tables
• Offer context — company, title, tech stack, scores embedded as comments
• Placeholders — for the AI agent to fill from your cv-master.md

The agent then fills the placeholders and you convert to PDF:

applyr cv pdf ~/.applyr/cv/cv-acme-ai-engineer.html

The PDF is generated with Chrome headless, no headers or footers.

---

All commands

Command	Description
applyr init	Set up ~/.applyr/ (config, database, agent instructions)
applyr add '<json>'	Register a new job offer
applyr list [--status S] [--sort F]	List offers (default: last 50)
applyr pipeline [--min-score N]	View offers grouped by status
applyr show <id>	Show full offer details with topic scores
applyr update <id> <status> [--notes ""]	Update offer status
applyr delete <id>	Delete an offer
applyr search <keyword> [--status S]	Search by company/title/notes/tech
applyr stats	Conversion funnel, channels, salary, work mode
applyr gaps [--limit N]	Skill gap analysis by frequency
applyr followups	Pending/overdue follow-ups with contact info
applyr trends [--period week|month]	Application trends over time
applyr summary [--json]	Weekly summary (JSON for LLM consumption)
applyr export [--format csv|json]	Export all data
applyr cv generate <id>	Generate ATS-safe HTML CV skeleton
applyr cv pdf <file.html> [--output f.pdf]	HTML to PDF via Chrome
applyr version	Show version
applyr help	Show help

---

Offer fields reference

Field	Type	Valid values	Required
title	string	Any	Yes
company	string	Any	No
summary	string	Any	No
date_received	string	YYYY-MM-DD	No
date_applied	string	YYYY-MM-DD	No
status	string	pending, applied, waiting, in_process, rejected, discarded, offer	No
canal	string	linkedin_easy, linkedin_direct, email, portal, referral, other	No
work_mode	string	remote, hybrid, onsite	No
location	string	Any	No
salary_min	integer	Annual EUR	No
salary_max	integer	Annual EUR	No
salary_period	string	annual, monthly	No
seniority_level	string	junior, mid, senior, lead, director	No
role_category	string	backend, frontend, fullstack, ai, devops, data, mobile, qa, other	No
tech_stack	string	Comma-separated	No
cover_letter	integer	0 or 1	No
cover_letter_file	string	File path	No
contact_name	string	Any	No
contact_role	string	Any	No
job_url	string	URL	No
rejection_reason	string	Any	No
notes	string	Any	No
topics	object	See Scoring section	No

---

Scoring

When you provide topics in applyr add, the compatibility score is auto-calculated using weighted averages:

Topic	Default Weight	What to evaluate
tech_stack	30%	How much of the required tech does the user know?
education	15%	Does the education match what they ask?
english	10%	Does the language level meet the requirement?
experience	15%	Years, seniority, and industry match?
projects	20%	Are the user's projects relevant to this role?
cultural_fit	10%	Work mode, company culture, location match?

Each topic score goes from 0 to 100. The weighted average becomes the compatibility percentage.

Default threshold to recommend applying: 65% (configurable).

Customize weights, topic names, and threshold in ~/.applyr/applyr.toml.

---

Status flow

pending ──> applied ──> waiting ──> in_process ──> offer
               |            |           |
               v            v           v
           discarded    rejected    rejected

• pending — offer registered, not yet applied
• applied — application sent (auto-schedules follow-up)
• waiting — waiting for company response
• in_process — interview stage
• offer — offer received
• discarded — decided not to apply
• rejected — company rejected your application

---

Configuration

Edit ~/.applyr/applyr.toml:

[general]
threshold = 65          # Min compatibility % to recommend applying
followup_days = 10      # Days before follow-up reminder

[weights]
# Relative importance of each topic (auto-normalized, no need to sum to 1.0)
tech_stack = 30
education = 15
experience = 15
projects = 20
english = 10
cultural_fit = 10

[cv]
# cv_master = "~/.applyr/cv-master.md"
# output_dir = "~/.applyr/cv"

---

Data storage

All data is stored locally in ~/.applyr/jobs.db (SQLite). Nothing leaves your machine.

Export anytime:

applyr export --format json --file my-applications.json
applyr export --format csv --file my-applications.csv

---

Example conversation with your AI agent

You:   "Analyze this job posting for AI Engineer at Acme"
Agent: Reads the posting + your cv-master.md
       Evaluates compatibility per topic
       Runs: applyr add '<json with all fields>'
       → "Registered as #42 — 78% match. Gaps: English, Experience"

You:   "Apply to it"
Agent: Runs: applyr update 42 applied --canal linkedin_easy
       Runs: applyr cv generate 42
       Fills placeholders from cv-master.md
       Runs: applyr cv pdf ~/.applyr/cv/cv-acme-ai-engineer.html
       → "CV generated. PDF ready at ~/.applyr/cv/cv-acme-ai-engineer.pdf"

You:   "What skills should I focus on improving?"
Agent: Runs: applyr gaps
       → "Experience appears in 15 offers (avg gap 20%). English in 12 offers."

You:   "How am I doing this month?"
Agent: Runs: applyr summary --json
       → Structured JSON with applications sent, response rate, trends

You:   "Any follow-ups due?"
Agent: Runs: applyr followups
       → "3 overdue: Acme (#42, 5 days ago), Beta (#38, 3 days ago)..."

---

License

MIT