career-ops

Product Introduction

1. Definition: career-ops is an open-source, AI-powered job search automation system designed as a local-first command-line interface (CLI) tool. It functions as a collection of prompts or "skills" that run within existing AI coding assistants like Claude Code, GitHub Copilot CLI, or OpenCode.
2. Core Value Proposition: It exists to automate and privatize the technical job search process. It solves the inefficiency of manual job hunting by using AI to intelligently scan, evaluate, and prepare applications, all while keeping sensitive user data—CVs, profiles, application history—entirely on the user's local machine with zero telemetry.

Main Features

1. AI-Powered Job Scoring & Rubric Evaluation: The system uses a configured Large Language Model (LLM) to evaluate job descriptions against a user's CV. It employs a published five-dimension rubric (match, north-star alignment, compensation, cultural signals, red flags) to generate a holistic score from 1.0 to 5.0, complete with citations to specific CV lines and JD requirements. This moves beyond simple keyword matching to a nuanced, criteria-based assessment.
2. Local-First Portal Scanning (Zero-Token): career-ops includes pre-configured web scrapers for over 150 company career portals (Greenhouse, Ashby, Lever). The /career-ops scan command runs these scrapers locally, fetching job listings without using paid third-party API tokens. The resulting list is then processed and ranked by the AI scoring engine.
3. Tailored Application Material Generation: For roles a user chooses to apply for, the tool automates busywork. It generates ATS-optimized, tailored PDF resumes for each specific role. Furthermore, it can read open-ended application questions on platforms like Greenhouse and draft context-aware answers based on the job description and the user's CV, providing paste-ready text for the user to review and edit.
4. Terminal-Based Pipeline Dashboard: The entire job search pipeline is managed and tracked through a Go-based terminal dashboard. This provides a centralized, private view of scanned jobs, scores, application statuses, and generated materials, replacing manual spreadsheets.

Problems Solved

1. Pain Point: The manual, repetitive, and time-consuming nature of modern job searching—scrolling portals, tailoring resumes, answering similar application questions—is a significant drain for technical professionals.
2. Target Audience: The primary user personas are software developers, data scientists, machine learning engineers, DevOps engineers, and other technical professionals actively seeking new roles. It is especially valuable for senior and lead-level candidates who require highly tailored applications and have complex career narratives to align.
3. Use Cases: A developer beginning a job search can use /career-ops scan to pull hundreds of relevant listings in minutes instead of hours. A candidate applying to 20 roles can use the tool to generate 20 uniquely tailored resumes and draft answers to common essay questions, saving dozens of hours. A privacy-conscious individual can manage their entire search pipeline offline, avoiding cloud-based SaaS platforms.

Unique Advantages

1. Differentiation: Unlike cloud-based resume checkers (e.g., Jobscan) or auto-apply tools, career-ops is open-source (MIT-licensed), runs locally, and maintains a human-in-the-loop philosophy—it never auto-submits applications. Unlike recruiter platforms, it has no account, no subscription fee for the tool itself, and no data sharing.
2. Key Innovation: Its architecture is AI-CLI-agnostic. It is not a standalone SaaS or a locked-in skill for one provider. It is built on the Open Agent Skill Standard, allowing its prompt-based "modes" to run inside any compatible AI coding CLI (Claude Code, OpenCode, etc.). This decouples the tool's logic from the AI provider, giving users choice and future-proofing.

Frequently Asked Questions (FAQ)

1. Is career-ops completely free to use? Yes, the career-ops software is permanently free, open-source, and MIT-licensed. The only potential cost is the subscription fee for the AI coding CLI you choose to run it with (e.g., Claude Pro, GitHub Copilot). The project is sustained through voluntary community patronage via GitHub Sponsors.
2. How does career-ops protect my privacy and data? All user data—CV, profile, application pipeline, and generated documents—resides exclusively in plain text files (Markdown, YAML) on your local machine. The system has no telemetry, no accounts, and does not upload your data to any career-ops server. Data only leaves your machine in prompts sent to your chosen AI CLI provider, governed by their privacy policy.
3. Can career-ops automatically submit job applications for me? No, and this is a deliberate design choice. career-ops automates the preparation (scanning, scoring, resume tailoring, answer drafting) but stops short of the final submission click. This ensures the user maintains quality control, prevents "spray-and-pray" tactics that harm recruiter relationships, and avoids potential ATS blacklisting from automated submissions.
4. What AI models or coding CLIs are required to run career-ops? career-ops works as a "skill" within several AI coding CLI environments. First-class supported CLIs include Claude Code, OpenCode, Antigravity CLI, Grok Build CLI, and GitHub Copilot CLI. You must have one of these CLIs installed and configured for career-ops to function, as it provides the underlying LLM execution engine.
5. How accurate and reliable is the AI job scoring rubric? The scoring is based on a transparent, published methodology. It uses an LLM guided by a structured rubric, not a simple formula, aiming for nuanced judgment similar to a human screener. However, as with any AI evaluation, its output is advisory. The final decision on whether a 4.2-scoring role is a "good fit" remains with the user, who is expected to review the AI's citations and reasoning.