career-ops: AI Job Search Command Center
Turn any AI coding CLI into a complete job search pipeline: evaluate offers, generate tailored resumes, scan portals, track applications — all running locally.
Evidence shows: README states local operation, data sent only to user-chosen AI provider, and default prohibition of auto-submission with human confirmation. However, no code-level evidence for least privilege, and dependencies like Playwright may have broad permissions. Deductions: no actual code audit shown, dependency security only based on package.json versions without vulnerability verification.
Evidence shows: clear project structure, consistent script naming, test fixtures and CI workflows. But no specific failure message examples, and dependency availability not verified. Deductions: insufficient evidence for failure messages, dependency availability only asserted.
Evidence shows: support for multiple CLIs and models, various usage scenarios and configuration options. But capability boundaries not clearly defined, trigger precision depends on user input. Deductions: capability boundaries insufficiently described, trigger precision not detailed.
Evidence shows: multilingual README, installation instructions, FAQ, MIT license. But version only in package.json, no CHANGELOG, maintenance responsibility unclear. Deductions: missing version changelog, maintenance responsibility not explicit.
Evidence shows: outputs are PDFs and reports, marginal value present, cost-benefit reasonable. But no actual run results. Deductions: output usability not verified, marginal value based on claims.
Evidence shows: usage cases and statistics in README, but no traceable verification data. Deductions: claims lack traceability, cross-source corroboration insufficient, facts and inferences not clearly separated.
- Dependencies like Playwright may have broad permissions; actual code needs review.
- No changelog provided; maintenance responsibility unclear.
- Claims lack traceability; independent verification needed.
What does this agent do, and when should you use it?
career-ops (career-ops.org) is an open-source, CLI-agnostic job-search command center that turns AI coding CLIs like Claude Code, Codex, OpenCode, and Antigravity into automated job search pipelines. It evaluates job listings with a structured A-G evaluation framework, generates ATS-optimized PDF resumes and cover letters, scans job portals (Greenhouse, Ashby, Lever, etc.) using Playwright, and maintains a centralized application tracker. The tool includes legitimacy checks for postings and ghost-job detection, but is emphatically not an auto-applier — it always leaves the final decision to the user. The author used it to evaluate 740+ job offers, generate 100+ tailored CVs, and land a Head of Applied AI role. It is MIT-licensed and follows the CareerOps Manifesto.
career-ops runs inside an AI coding CLI, reading your CV (cv.md) and configuration (config/profile.yml) and executing tasks via modes (modes/*.md). When you paste a job URL or description, it detects the role type, performs an A-G evaluation (role summary, CV match, level strategy, compensation research, personalization, interview prep, and legitimacy check), generates reports (reports/), ATS-optimized PDFs (output/), and updates the tracker (data/pipeline.md). It offers commands via modes (scan, batch, pdf, cover, email, contacto, deep, etc.), supports 100+ pre-configured companies, and uses Playwright for portal scanning. The dashboard (dashboard/) is a Go + Bubble Tea TUI for browsing and sorting applications. After installation, users configure it by chatting with the CLI, then paste job links into the session; the AI handles the rest. All data remains local, and control stays with the user.
- A job seeker overwhelmed by hundreds of listings wants to filter out ghost jobs and focus on the best fits: use the A-G evaluation and legitimacy checks to score and prioritize offers.
- A candidate needs a tailored resume and cover letter for each application: use scan, pdf, and cover modes to automatically generate ATS-optimized documents.
- A developer wants to automate job search without installing extra tools: run career-ops in an existing AI coding CLI, leveraging their Node.js environment.
- A product-minded engineer wants to build a custom job search pipeline: modify modes, scoring weights, and templates to suit their exact needs.
- A privacy-conscious user wants to control their data: career-ops keeps all data local, only sending necessary context to the chosen AI provider.
- A job seeker wants to track applications and analyze rejection patterns: use followup-cadence, stats, and analyze-patterns to refine strategy.
What are this agent's strengths and limitations?
- Truly CLI-agnostic: runs as a skill in Claude Code, Codex, OpenCode, Antigravity, and others, avoiding vendor lock-in.
- Structured A-G evaluation framework with explicit scoring and legitimacy checks to avoid ghost jobs and scams.
- Automates generation of ATS-optimized PDF resumes and cover letters, saving time.
- Pre-configured with 100+ company portals and customizable search queries.
- Human-in-the-loop design ensures AI never auto-submits applications, keeping control with the user.
- Local data storage preserves privacy and data ownership.
- Initial evaluations may be poor; requires significant user investment to provide context (CV, career story, preferences).
- Requires Node.js and Playwright installation, adding setup steps.
- Portal scanning depends on ATS APIs and web scraping, which may be unstable.
- TUI dashboard is experimental (alpha) and may contain bugs.
- Does not support auto-apply for some ATS systems (e.g., Workday, LinkedIn) without additional scripting.
- Requires active maintenance: users need to regularly run scans and updates.
How do you install or deploy this agent?
Prerequisites: Node.js (bundled with most AI CLIs). Quick start with npx: run npx @santifer/career-ops init. This clones the latest release and installs dependencies. Alternatively, use git clone https://github.com/santifer/career-ops.git && cd career-ops && npm install, then set up Playwright with npx playwright install chromium (only for PDF generation). Run npm run doctor to validate prerequisites, copy the template config (cp config/profile.example.yml config/profile.yml), and create your cv.md in the project root. For a global install, run npm i -g @santifer/career-ops to get the career-ops command. Customize portals.yml with your target companies.
How do you use this agent?
After installation, open your AI coding CLI in the project directory (e.g., claude, codex, or opencode). On first launch, chat with the CLI to set up your CV, profile, and target roles. To trigger the pipeline, paste a job URL or description into the session; career-ops auto-detects and runs the full flow. Use slash commands: /career-ops scan to scan portals, /career-ops pdf to generate resumes, /career-ops batch for parallel evaluation, /career-ops tracker to view statuses. For headless batch workers, use claude -p or codex exec. Launch the dashboard TUI with npm run serve:dashboard. Always review outputs before submitting.