career-ops
Santiago Fernández de Valderrama (santifer) & community · Open-Source Job Search Agent That Runs in Your AI Coding CLI
career-ops is a free, MIT-licensed job search system that runs locally inside AI coding CLIs like Claude Code, Codex, OpenCode, or Gemini CLI to scan job boards, score listings against your CV, generate tailored PDF resumes, and draft application answers. It is for job seekers comfortable with a terminal who want help triaging and preparing applications while submitting every one themselves.
Best for
Technical job seekers who already use an AI coding CLI and want a local, private system to triage listings and tailor applications while keeping final control
Not ideal for
Non-technical job seekers who want a point-and-click web app, anyone looking to auto-apply at volume, or users who want a hosted service with no local setup
Who it's for
Job seekers, especially software and technical professionals, who work in AI coding CLIs and want local, reviewable job search automation
Capabilities
- Runs inside Claude Code, Codex, OpenCode, Antigravity CLI, GitHub Copilot CLI, Gemini CLI, Qwen Code, and other agent-skill-compatible CLIs
- Scans pre-configured company career pages and ATS job boards (Greenhouse, Lever, Ashby, Workday, and more) without spending model tokens on the scan
- Structured evaluation report per listing with a 1–5 score across match, target-role fit, compensation, cultural signals, and red flags
- Separate legitimacy assessment (High Confidence, Proceed with Caution, Suspicious) to flag scam and ghost postings
- Generates ATS-optimized PDF resumes and cover letters tailored to each role
- Drafts answers to application form questions for you to review and submit manually
- Local application tracker with funnel and rejection-pattern stats plus a Go terminal dashboard
- Batch evaluation across multiple CLIs, added in v1.33.0
Limitations
- career-ops does not provide a model: it needs an AI coding CLI, and while a free tier (Antigravity CLI) or local models can work, evaluation and CV quality depend on the model you use
- Setup needs Node.js 18+ (22.5+ recommended), Git, and Playwright's headless Chromium, which assumes terminal comfort
- The rule against fabricating skills or experience is enforced only in prompts, not by code, so every generated CV and answer must be reviewed
- No hosted service: everything runs on your machine, and the optional local web UI is still experimental (alpha)
- Users are responsible for complying with job sites' terms of service; the project warns against spamming employers or ATS systems
- If an ANTHROPIC_API_KEY is set, it overrides Claude Pro/Max subscription billing, which can lead to unexpected API charges
Use cases
- Scanning dozens of target companies' career pages each morning and shortlisting roles that score 4.0 or higher
- Pasting a job URL to get a fit report, compensation research, and a tailored PDF resume in one run
- Comparing several offers or listings side by side with a ranked summary
- Drafting answers to Greenhouse, Lever, or Ashby application questions before reviewing and submitting them yourself
- Spotting rejection patterns by ATS channel in the local tracker to adjust where you apply
career-ops pricing
Open source
Free
- MIT license
- No account, waitlist, or telemetry
- Runs locally in your AI coding CLI
Note: career-ops is free and community-funded. Any costs come from the AI coding CLI subscription or model API usage you run it with; the project documents a zero-cost setup on Antigravity CLI's free tier.
Technical specs
Frequently asked questions
What is career-ops?
career-ops is an open-source job search system that runs in your AI coding CLI. It scans job sources, scores listings against your CV and goals, generates tailored resumes and cover letters, drafts application answers, and tracks your pipeline locally.
Is career-ops free?
Yes. career-ops is MIT-licensed with no paid tier or account. Any cost comes from the AI coding CLI or model API you run it with; the project documents a zero-cost setup on Antigravity CLI's free tier and options for cheaper or local models.
Does career-ops apply to jobs automatically?
No. It drafts answers for every form field and prepares documents, but it never submits an application or sends an email. You review everything and submit it yourself.
Which AI tools does career-ops work with?
Its docs list Claude Code, Codex, OpenCode, Qwen Code, GitHub Copilot CLI, and Gemini CLI, the README adds native Antigravity CLI and Grok Build CLI support, and the project describes itself as CLI-agnostic for agents that support its skill format.
Where is my data stored?
Your CV, profile, pipeline, and reports are kept as local Markdown and YAML files on your machine and are sent only to the AI provider you choose. The project states it does not collect or store your data.
Integrations & fit
Similar agents
Microsoft
Individuals and teams using Microsoft 365 and Windows who want AI inside the apps they already use
FreemiumNotion
Teams already using Notion who want AI agents and writing assistance embedded in their existing docs, databases, and meetings
PaidJasper AI
Marketing teams producing high volumes of content who need consistent brand voice and agent-driven workflows
PaidAbout career-ops
career-ops is not a hosted app; it is a set of agent skills and scripts you install into a local folder with one npx command and then drive from whichever AI coding CLI you already use. After a short onboarding conversation about your CV, target roles, compensation, and location, it scans pre-configured company career pages and applicant tracking systems such as Greenhouse, Lever, Ashby, and Workday, then turns each promising listing into a structured report with a 1–5 score. The published methodology is deliberately a holistic model judgment across five dimensions — skills match, fit with your target roles, compensation against market data, cultural signals, and red flags — with 4.0 as the suggested apply line and a separate legitimacy check that flags likely scam or ghost postings. For roles worth pursuing it generates an ATS-friendly PDF resume, cover letter, and drafted answers to application form fields, and logs everything to a local tracker with a terminal dashboard. The human-in-the-loop design is explicit: it never submits an application or sends an email. Everything stays in local Markdown and YAML files, sent only to the model provider you choose. The costs and caveats are practical ones: you need an AI coding CLI plus Node.js, Git, and Playwright; the model is the main cost, though the project documents a no-cost path on Antigravity CLI's free tier and routing to cheaper or local models, with results depending on the model you pick; the rule against fabricating experience is enforced only through prompts, and you are responsible for reviewing every document and respecting job sites' terms of service.
Updates from career-ops
career-ops v1.33.0 introduced batch evaluation across multiple AI CLIs, fetching job descriptions from the Greenhouse, Lever, and Ashby APIs, and new providers including the Gem job board API and Telegram channels.
v1.32.0 added discard-log analytics and tuning summaries, an H1B sponsorship history plugin, the Built In aggregator as a provider, and durable scan receipts.
Are you the founder? Claim this listing →