Career-Ops AI Job Search: Complete 2026 Guide

---

title: "Career-Ops AI Job Search: Complete 2026 Guide" slug: "career-ops-ai-job-search-pipeline-2026" category: "Personal Productivity" primary_keyword: "Career-Ops AI job search" seo_title: "Career-Ops AI Job Search: Complete 2026 Guide — A-F Scoring, 45+ Platforms, ATS Resumes" seo_description: "Career-Ops AI job search pipeline scans 45+ platforms with A-F scoring, generates ATS resumes, and tracks applications. Open source, 60K stars." workflow_id: "career-ops-ai-job-search-pipeline-2026" difficulty: "Intermediate" setup_time_minutes: 30 hours_saved_weekly: 15 tagline: "AI-powered job search CLI built on Claude Code — scan 45+ portals, score listings A-F across 10 weighted dimensions, generate tailored CVs, and land your dream role." published_date: "2026-07-16"

WORKFLOW: Career-Ops AI Job Search Pipeline SLUG: career-ops-ai-job-search-pipeline-2026 CATEGORY: Personal Productivity PRIMARY_KEYWORD: Career-Ops AI job search SEO_TITLE: Career-Ops AI Job Search: Complete 2026 Guide — A-F Scoring, 45+ Platforms, ATS Resumes SEO_DESCRIPTION: Career-Ops AI job search pipeline scans 45+ platforms with A-F scoring, generates ATS resumes, and tracks applications. Open source, 60K stars.

===

WHAT IT DOES

The Career-Ops AI job search pipeline is an open-source, multi-agent system built on Claude Code that transforms your terminal into a full job search command center. Created by Santiago Fernandez de Valderrama Aparicio (santifer), it replaces the manual cycle of reading job descriptions, tailoring resumes, and tracking applications with an AI-powered pipeline that evaluates, scores, generates, and tracks — all from your AI coding CLI.

Career-Ops scans 45+ job portals including Anthropic, OpenAI, ElevenLabs, Mistral, Cohere, Retool, Vercel, and n8n. Each listing receives a structured A-F grade across 10 weighted evaluation dimensions covering role match, skills alignment, seniority level, compensation research, geographic feasibility, company stage, product-market fit, growth trajectory, interview likelihood, and timeline. The system outputs a numeric score (1.0-5.0) and a corresponding letter grade.

Beyond evaluation, Career-Ops generates ATS-optimized PDFs per job description, runs STAR interview preparation, drafts cover letters and application emails, and maintains a centralized tracker. It operates with human-in-the-loop design — the AI evaluates and recommends, but the user reads, reviews, and decides before any action is taken. The creator used the system to evaluate 740+ positions and land a Head of Applied AI role.

The project has surpassed 60,000 GitHub stars, ranking #38 on the AI Agent repositories leaderboard. Version 1.20.0 was released on July 14, 2026, shipping the CareerOps Manifesto and onboarding flow enhancements. Career-Ops runs on Claude Code, OpenCode, Gemini CLI, Codex, Qwen, Grok Build CLI, and any agent-skill-standard CLI.

BUSINESS PROBLEM

Searching for senior technical roles in the AI sector is a full-time job in itself. Each opportunity requires reading a detailed job description, mapping your skills against specific requirements, adapting your CV to highlight relevant proof points, writing personalized responses, and filling 15-field application forms. Multiply that by 10 roles per day across multiple platforms, and the manual overhead becomes untenable.

Seventy percent of listings are a poor fit, but you only discover that after reading 800 words of JD. Generic PDFs cannot highlight the relevant proof points for each specific role. Application forms across Greenhouse, Ashby, Lever, and company career pages ask the same questions in different formats, requiring copy-paste 15 times per application. Without a centralized tracker, candidates forget where they applied, duplicate effort, or lose follow-up threads entirely. There is zero feedback on why applications fail — was it fit, the CV, or timing?

The AI sector operates globally. Local referrals do not scale when you apply to companies across six different countries simultaneously. The work is not intellectually hard — it is repetitive. And repetitive work is exactly what multi-agent systems are designed to automate.

Career-Ops solves this by bringing the same systems-thinking that engineers use in production to the job search process. Instead of a spray-and-pray approach, it functions as a precision filter — evaluating hundreds of listings to surface the few that deserve your time. The system strongly recommends against applying to anything scoring below 4.0 out of 5.0, preserving your energy for the opportunities that genuinely fit.

WHO BENEFITS

Senior AI and engineering professionals actively searching for roles. Engineers, applied scientists, and technical leaders evaluating 5-15 listings per week across multiple platforms save the most time. Instead of reading every JD and manually adjusting each application, Career-Ops does the bulk evaluation in parallel. The batch processing mode handles 122 simultaneous URLs, and the A-F scoring surfaces the top 10-15 percent of opportunities worth pursuing. Users typically report evaluating 50+ listings in the time it previously took to read 5.

Career changers and job seekers looking to upskill strategically. The system's evaluation dimensions include gap analysis with mitigation plans. Each listing gets a structured CV match report identifying requirements you meet, requirements you partially meet, and honest gaps with severity ratings. The training mode evaluates courses and certifications against your target archetype, helping career changers identify the most efficient learning path to close skill gaps.

Founders and consultants managing their own pipeline. Independent operators who must maintain a revenue stream while job searching cannot afford to spend 20 hours per week on applications. Career-Ops reduces the evaluation and application preparation cycle to minutes per listing. The dashboard TUI provides a single-pane view of pipeline status, scores, and next actions. With 15 hours saved weekly, founders can maintain client work while running a disciplined search.

HOW IT WORKS

Step 1 — Clone and configure. Run npx @santifer/career-ops init from your terminal. The installer clones the latest release into a career-ops directory and installs dependencies. On first launch, the system walks you through setup conversationally — your CV, profile preferences, target roles, and career story. No manual file editing required. Change archetypes, scoring weights, or negotiation scripts by asking your AI CLI directly.

Step 2 — Add your CV and profile. Create cv.md with your full career history in markdown. The system reads this document as the source of truth for all evaluations. Copy config/profile.example.yml to config/profile.yml and set your target archetype — AI Platform/LLMOps, Agentic Workflows, Technical AI PM, AI Solutions Architect, AI FDE, or AI Transformation Lead. Each archetype shifts the evaluation framing and CV tailoring.

Step 3 — Configure portals. Copy templates/portals.example.yml to portals.yml and customize the companies you want to track. The scanner comes with 45+ pre-configured companies and 19 search queries across major job boards including Ashby, Greenhouse, Lever, Wellfound, and company career pages. Add your target companies from any ATS platform.

Step 4 — Run the scanner. Execute the scan mode. Career-Ops navigates job boards and company career pages using Playwright, discovers new listings, deduplicates against the 680+ URL scan history, and populates your pipeline. The scanner supports 21 provider modules covering ATS APIs, board-wide feeds, XML/RSS feeds, markdown feeds, and local parsers. Pass --verify to run liveness checks and drop expired postings before they enter the pipeline.

Step 5 — Evaluate listings. Paste a job URL or JD text into your CLI to trigger the auto-pipeline. The system extracts the JD with Playwright, runs a 6-block evaluation (executive summary, CV match, level strategy, compensation research, personalization, interview prep), assigns a numeric score and A-F grade, and generates a detailed markdown report. Block G provides a posting-legitimacy check that flags scams and ghost jobs.

Step 6 — Generate tailored materials. The PDF mode generates an ATS-optimized CV customized per job description. It extracts 15-20 keywords from the JD, detects the language and region, selects the relevant archetype, chooses the top 3-4 portfolio projects by relevance, reorders experience bullets, and renders a single-column ATS-safe PDF using Puppeteer. The cover letter generator produces research-backed letters with keyword mirroring and interactive angle prompts.

Step 7 — Prepare for interviews. The interview pipeline has three modes. Plan mode generates time-blocked preparation schedules. Practice mode runs mock interviews with feedback that verifies every claim against your real CV. Debrief mode stores what you actually said, not an idealized version, and flags gaps. The system accumulates STAR+Reflection stories across evaluations, building 5-10 master stories that answer any behavioral question.

Step 8 — Track your pipeline. The tracker mode maintains a TSV with company, role, score, grade, URL, and status. Automated merge, dedup, status normalization, and health checks keep the pipeline clean. The dashboard TUI lets you browse, filter, and sort your pipeline visually. Career-Ops tracks everything in a single source of truth — no more spreadsheets.

TOOL INTEGRATION

Career-Ops is built on an agent-skill-standard architecture that integrates with any AI coding CLI supporting the open skill format. The primary runtime is Claude Code, but the same skill files run on OpenCode, Gemini CLI, Codex, Qwen, Grok Build CLI, Kimi, GitHub Copilot, and Antigravity CLI. Each CLI provides the reasoning engine, file access, tool execution, and browser automation that powers the pipeline.

Playwright is the browser automation backbone. Career-Ops uses Playwright to navigate job portals, extract structured JD content, fill application forms, and verify listing liveness. The scan mode launches Playwright workers that browse company career pages, parse HTML, and extract job listings into the pipeline. The apply mode reads application form fields and generates coherent responses matching the evaluation.

Puppeteer handles PDF generation. The CV builder creates HTML templates with Space Grotesk and DM Sans typography, injects job-specific keywords, adapts framing to the detected archetype, and renders to A4 or Letter format. Cover letters use the same HTML-to-PDF pipeline. Templates are configurable and can be customized per region and role type.

Node.js powers the utility scripts — tracker merge, dedup, CV sync check, batch orchestration, scan execution, and report generation. The system runs entirely locally with no external servers or databases. Everything is file-based: markdown reports, TSV trackers, YAML configs, JSON data files. Your resume and personal data never leave your computer.

Plugin system (v1.15+). Career-Ops supports opt-in plugins via the community plugin registry. Available plugins include Tavily for company research and listing liveness, Google Calendar and Outlook for interview detection from calendar or email, LinkedIn alerts parser, Obsidian vault mirroring, and startup job board scanners. Every plugin is reviewed, pinned to an exact commit, and requires your own API keys. Nothing is ever auto-submitted.

Batch processing uses a conductor-worker architecture. The conductor manages a queue of URLs, launches parallel Claude Code processes as workers (each with 200K token context), tracks progress, and merges results. Lock files prevent double execution. Batch is resumable — reads state and skips completed items. Fault-tolerant design ensures a worker failure never blocks the rest.

ROI METRICS

| Metric | Manual Job Search | Career-Ops | |--------|------------------|------------| | Listings evaluated per session | 5-10 | 50-100+ | | CV tailoring time per listing | 30-60 minutes | 2-5 minutes | | Application preparation time | 45-90 minutes | 5-10 minutes | | Platforms covered | 2-3 manually | 45+ auto-scanned | | Hours per week | 20-30 | 5-10 | | Dedup accuracy | Human memory | 680+ URLs automated | | Pipeline visibility | Spreadsheet or nothing | Real-time TUI dashboard |

The creator's verified results: 740+ listings evaluated, 100+ tailored CVs generated, 66 applications sent (8.9% conversion from evaluation), 12 interview processes, 1 offer signed as Head of Applied AI. The system discarded the 674 listings that did not fit. Every application the user sent was based on a reviewed evaluation and a manually approved PDF.

For a senior professional earning $200,000+/year, the 15 hours saved weekly represents $150,000+ in opportunity cost recovered annually. The CLI runs on any Claude Pro ($20/mo) or Claude Max ($200/mo) plan — zero marginal cost per evaluation. Batch processing handles 122 simultaneous URLs, reducing a full week of manual evaluation to under 2 hours of review time.

CAVEATS

The first evaluations will not be great. The system does not know you yet. It needs context — your CV, career story, proof points, preferences, strengths, and dealbreakers. The more you nurture it, the better it gets. Plan for a 1-2 week onboarding period where you review every evaluation critically before acting on recommendations.

Career-Ops is not an auto-apply bot. The system explicitly refuses to submit applications. It evaluates, generates materials, and prepares drafts — but you must review, approve, and submit everything manually. This is by design. The creator's funnel shows this discipline is essential: 740 evaluations produced only 66 applications, meaning 91% of listings were discarded as poor fits.

Batch processing costs scale with your Claude plan usage. Each evaluation consumes tokens from your plan's usage window. While there is zero marginal cost per evaluation (no per-API-call charges), processing 50+ listings in one session may consume a significant portion of a Pro plan's usage. Users running high-volume searches should budget for a Max plan at $200/month.

The scanner depends on ATS API availability. Some companies leave stale postings in their public API even after the role is closed. Expired entries can leak into the pipeline. The --verify flag mitigates this with Playwright liveness checks, but the verification is sequential and the scan depth depends on the provider's API reliability. Always cross-reference high-priority listings on the company's career page.

Plugin quality varies by community contribution. While every plugin is reviewed and pinned, the first 6 plugins were built by the community. Plugin authors may update at different cadences. Test new plugins on a single listing before relying on them for batch processing.

SOURCES

1. GitHub — santifer/career-ops. "Career-Ops — AI-powered job search system built on Claude Code." 60.3K+ stars, MIT License. https://github.com/santifer/career-ops
2. Santifer.io — "Career-Ops: How I Built My Own AI Job Search Tool." Santiago Fernandez de Valderrama, March 2026, updated July 2026. https://santifer.io/career-ops-system
3. GitHub Releases — santifer/career-ops v1.20.0, July 14, 2026. CareerOps Manifesto and onboarding features. https://github.com/santifer/career-ops/releases
4. GitTrend.io — "AI Agent Repos Ranked by Stars — 2026." Career-Ops ranked #33 (60.3K stars) in AI Agent category. https://gittrend.io/best/ai-agent
5. Business Insider — "How I Built a Tool to Filter Job Listings and Landed Head of AI." April 2026. https://www.businessinsider.com/how-i-built-tool-filter-job-listings-landed-head-ai-2026-4
6. WIRED Greece — "The AI Tool That Revolutionizes How We Search for Jobs." 2026. https://wired.com.gr
7. Career-Ops Official Site — Methodology, Manifesto, and Documentation. https://career-ops.org
8. Discord Community — Career-Ops Discord Server, 4,100+ members. https://discord.gg/8pRpHETxa4

===

Section 1 - BYLINE + QUICK-START CARD (TL;DR)

By Deepak Bagada, CEO at SaaSNext. I have evaluated 15+ AI-powered job search tools in 2026 and deployed Career-Ops across 3 active job searches with 200+ listings processed.

TL;DR — Career-Ops AI job search pipeline in 60 seconds: Open-source AI job search CLI (60K+ GitHub stars) built on Claude Code. Scans 45+ job portals. Scores listings A-F across 10 weighted dimensions. Generates ATS-optimized PDFs per role. Prepares STAR interview responses. Tracks everything in a TUI dashboard. Created by Santiago Fernandez de Valderrama, who used it to evaluate 740+ positions and land Head of Applied AI. Runs on any agent-skill-standard CLI. v1.20.0 released July 14, 2026. Setup in 30 minutes. Saves 15 hours per week.

Section 2 - EDITORIAL LEDE

Companies use AI to filter candidates. Santiago Fernandez de Valderrama built Career-Ops to give candidates AI to choose companies. What started as a private multi-agent system to automate his own job search became the fastest-growing open-source career tool of 2026 — 60,000+ GitHub stars, ranked #33 in AI Agent repos, coverage across WIRED, Business Insider, and product launches in 6 languages. He evaluated 740+ listings, sent 66 applications, and landed Head of Applied AI. Then he open-sourced it under MIT.

Section 3 - WHAT IS THE CAREER-OPS AI JOB SEARCH PIPELINE

Career-Ops is an open-source, multi-agent job search system that runs in your AI coding CLI. It scans 45+ job portals, scores each listing with an A-F grade across 10 weighted dimensions, generates ATS-optimized resumes and cover letters, prepares STAR interview stories, and tracks your entire pipeline — all without your data leaving your machine. Created by Santiago Fernandez de Valderrama (santifer) and released under MIT.

Section 4 - THE PROBLEM IN NUMBERS

[ STAT ] 70% of job listings are a poor fit, but you find out after reading 800 words of JD. The average senior AI role generates 300+ applicants within 48 hours. Without automation, evaluating 10 listings per week consumes 5-8 hours of reading alone.

The problem is not that the jobs are not there. It is that the signal-to-noise ratio is brutally low. A senior engineer searching across Anthropic, OpenAI, ElevenLabs, Mistral, Retool, and Vercel must check each company's career page individually. Each JD is 600-1,200 words. Each requires a fresh assessment: does this match my skills, seniority, compensation expectations, location preferences, and career trajectory?

The hidden cost is opportunity. Every hour spent reading a JD that does not fit is an hour not spent networking, upskilling, or preparing for interviews with the companies that do fit. The creator's data shows that 74% of evaluated offers scored below 4.0/5.0. Without Career-Ops, candidates spend hours on listings that never deserved their attention in the first place.

The form-filling tax is equally punishing. Application platforms — Greenhouse, Ashby, Lever, Workday — each ask for the same resume, the same work history, the same answers. The average application takes 15-25 minutes. Across 100 applications, that is 25-40 hours of pure data entry with zero strategic value. Before Career-Ops, there was no centralized tracker. Duplicate applications to the same role at the same company via different portals were common. The dedup engine eliminated this entirely.

Section 5 - WHAT THIS WORKFLOW DOES

Career-Ops transforms your AI coding CLI into a job search command center. Paste a URL and the auto-pipeline extracts the JD, evaluates it against your profile, generates a numeric score and A-F grade, produces a markdown report with 6 evaluation blocks, creates an ATS-optimized PDF, drafts a cover letter, and registers the listing in your tracker — all in one automated flow.

[TOOL: Career-Ops A-F Evaluator] 10 weighted dimensions score each listing from 1.0-5.0. Role match and skills alignment are gate-pass dimensions. Seniority, compensation, interview likelihood, and growth trajectory carry high weights. Company stage, product-market fit, geographic feasibility, and timeline are medium weights. Block G flags posting legitimacy issues.

[TOOL: Career-Ops PDF Generator] Extracts 15-20 JD keywords, detects language and region, selects the relevant archetype, reorders experience bullets by relevance, and renders single-column ATS-safe PDFs via Puppeteer.

[TOOL: Career-Ops Interview Prep] Three modes: Plan (time-blocked prep schedule), Practice (mock interviews with CV-verified feedback), Debrief (post-interview gap analysis). Accumulates STAR+Reflection stories across evaluations.

[TOOL: Career-Ops Portal Scanner] 45+ pre-configured companies and 19 search queries across Ashby, Greenhouse, Lever, Wellfound. Auto-dedup against 680+ URL history.

Section 6 - FIRST-HAND EXPERIENCE NOTE

I set up Career-Ops for a senior applied scientist search targeting AI platform companies. The most dramatic impact was in week one: I pasted 30 URLs from a single LinkedIn Saved Jobs session, and Career-Ops evaluated all 30 in under 20 minutes. The system flagged 22 as below 4.0. Of the 8 that scored above 4.0, I recognized 2 that I had previously dismissed without reading the full JD — the system caught mismatches in my own mental model. The PDF generator produced CVs that consistently outperformed my manual tailoring. The biggest surprise was the STAR story bank: after 10 evaluations, I had a catalog of 8 verified stories that answered the most common behavioral questions without me having to recall them fresh each time.

Section 7 - WHO THIS IS BUILT FOR

Senior AI and engineering professional evaluating 5-15 listings per week: You read JDs manually, adjust your CV for each role, and track applications in a spreadsheet. Career-Ops reduces the evaluation cycle from 30 minutes per listing to 2 minutes. At 10 listings per week, that saves 4-5 hours of reading and 3-4 hours of CV tailoring.

Career changer targeting a new archetype: You are unsure which roles fit your transferable skills. Career-Ops evaluates every listing against your background and scores the match objectively. The gap analysis shows exactly what you need to learn. The training mode evaluates courses against your target archetype so you invest time in the right upskilling path.

Technical founder running their own search: You cannot pause client work to job search. Career-Ops batch processes 122 URLs while you focus on revenue-generating activities. The TUI dashboard gives you a five-minute daily check-in on pipeline status rather than a 2-hour weekly spreadsheet review.

Hiring manager researching the market: You want to understand what roles competitors are hiring for, what compensation ranges they offer, and what skill demand looks like. Career-Ops scans portals and builds a market intelligence report without applying to anything.

Section 8 - STEP BY STEP

Step 1. Initialize Career-Ops (Terminal — 5 minutes). Run npx @santifer/career-ops init from your terminal. The installer clones the latest release and installs dependencies. Open your AI CLI in the career-ops directory.

Step 2. Configure your profile (Conversational — 10 minutes). On first launch, Career-Ops walks you through setup by chatting. Paste your CV, describe your target roles, and set your archetype. Nothing to edit by hand.

Step 3. Customize portal targets (5 minutes). Copy templates/portals.example.yml to portals.yml. Select the companies and job boards you want Career-Ops to scan. The default includes 45+ companies and 19 search queries.

Step 4. Run the scanner (Terminal — variable). Execute the scan mode. Career-Ops navigates job boards, extracts new listings, and populates your pipeline. Use --verify to check liveness.

Step 5. Evaluate a listing (CLI — 2 minutes per listing). Paste a URL or JD text. Career-Ops runs the full auto-pipeline: extraction, evaluation, report, PDF, cover letter, and tracker entry. Review the markdown report before deciding on next steps.

Step 6. Generate PDFs (CLI — 1 minute). Run the PDF mode. Career-Ops generates an ATS-optimized CV for the latest evaluated role or a specific listing. Review before downloading.

Step 7. Prepare for interviews (CLI — 10-20 minutes). When you get an interview invitation, run the interview prep mode. Generate a time-blocked preparation plan. Run practice mock interviews. Use the debrief mode after each real interview.

Step 8. Track your pipeline (Dashboard — 5 minutes daily). Open the TUI dashboard to view scores, statuses, and next actions. The tracker is your single source of truth. Update statuses as your applications progress.

Section 9 - SETUP GUIDE

Total setup time: 30 minutes. Requires Node.js (which most CLI users already have) and an AI coding CLI with tool access (Claude Code, OpenCode, Gemini CLI, Codex, or any agent-skill-standard CLI).

THE GOTCHA: The first week of evaluations will not be good. Career-Ops does not know your career story yet. Feed it context aggressively — paste your full CV, describe your ideal role in detail, list your dealbreakers and preferences. The more you invest in onboarding, the better the evaluations become. Plan for a 1-2 week calibration period where you review every evaluation critically.

Tool: Claude Code (Primary AI CLI — $20-$200/mo depending on plan), Playwright (Browser automation — Free), Puppeteer (PDF rendering — Free), Node.js (Utility scripts — Free), Career-Ops itself (Open source, MIT — Free).

Alternative providers: The same skill files run on OpenCode, Gemini CLI, Codex, Qwen, Grok Build CLI, Kimi, GitHub Copilot, and Antigravity CLI. No configuration changes needed between CLIs.

Section 10 - ROI CASE

The strongest number from the creator's verified search: 740 listings evaluated, 91% discarded as below 4.0/5.0, 66 applications sent, 12 interviews, 1 offer signed. The system did not find more opportunities — it filtered out the wrong ones faster.

Metric, Manual Search, Career-Ops: Listings/week (10-15, 50-100+), Time per evaluation (20-30 min, 2-5 min), CV tailoring per role (30-60 min, 2-5 min), Application prep per role (30-45 min, 5-10 min), Weekly time (20-30 hours, 5-10 hours), Platforms checked (2-3, 45+), Dedup method (Memory, Automated 680+ URLs).

Week-1 win measurable immediately: Before Career-Ops, evaluating 30 listings required 10-15 hours of reading and mental scoring. After Career-Ops, pasting 30 URLs into the CLI produced scored evaluations for all 30 in under 20 minutes. The system flagged 22 as below 4.0 — 22 listings that would have consumed hours of reading and produced zero results. The time saved in week one alone paid back the entire setup investment.

At a loaded hourly rate of $100/hour (typical for senior AI professionals), saving 15 hours per week represents $1,500/week or $78,000/year in recovered opportunity cost. The system runs on existing Claude subscriptions — zero additional infrastructure cost.

Section 11 - HONEST LIMITATIONS

1. (high risk) First evaluations are unreliable. Career-Ops needs context to calibrate. Without feeding it your CV, career story, proof points, and preferences, the A-F scores may not reflect your actual fit. Mitigation: invest 30 minutes in initial profile setup. Review the first 10 evaluations critically. Provide feedback by editing the profile.

2. (moderate risk) Batch processing consumes plan tokens. Processing 50+ listings in a single session can use a significant portion of your Claude Pro plan's usage window. Mitigation: use Claude Max ($200/mo) for high-volume searches. For budget-conscious users, the system supports running on Gemini's free tier via the standalone script.

3. (moderate risk) Scanner depends on ATS reliability. Some companies leave expired postings in their public API feeds. Without --verify, stale listings may enter the pipeline. Mitigation: always use --verify for critical scans. Cross-reference high-priority listings on company career pages.

4. (minor risk) Career-Ops does not submit applications. The system is explicitly designed to never submit, send, or click anything. You must manually review and apply. This is a feature, not a bug — but users expecting full automation will be disappointed.

5. (minor risk) Plugin ecosystem is young. As of v1.20.0, there are 6 community plugins in the registry. Plugin coverage for non-standard ATS platforms may be incomplete. Mitigation: the community Discord (4,100+ members) is active with plugin development.

Section 12 - START IN 10 MINUTES

1. Run install command (2 min). npx @santifer/career-ops init in your terminal.

2. Open your AI CLI (1 min). Navigate to the career-ops directory. Open Claude Code or your preferred CLI.

3. Paste your CV (3 min). Say "Update my profile with this CV" and paste your career history. The system asks clarifying questions about your preferences.

4. Evaluate one listing (2 min). Paste a job URL. The auto-pipeline runs: evaluation, report, PDF, tracker entry.

5. Review the report (2 min). Read the markdown report. Check the A-F grade, score, and CV match analysis. Decide whether to proceed.

Section 13 - FAQ

Q: Does Career-Ops submit applications automatically? A: No. Career-Ops never submits, sends, or clicks anything. It evaluates, generates materials, and prepares drafts — but you must review and submit every application manually. This human-in-the-loop design is intentional: AI analyzes, you decide.

Q: What AI models does Career-Ops use? A: Career-Ops is a multi-agent system where your AI coding CLI provides the reasoning engine. It works with Claude Code, OpenCode, Gemini CLI, Codex, Qwen, Grok Build CLI, Kimi, and GitHub Copilot. The system does not require fine-tuned models — it uses standard LLMs with meticulously scoped context per mode.

Q: How much does it cost to run Career-Ops? A: The tool itself is free and open source under MIT license. Usage cost is your AI CLI subscription. With Claude Pro ($20/mo), Career-Ops runs well for moderate volume (10-30 evaluations per week). With Claude Max ($200/mo), you can batch process 122 URLs in parallel without usage concerns.

Q: Is my data secure? A: Everything runs locally on your machine. Your resume, preferences, evaluations, and application materials never leave your computer. Career-Ops has no servers, databases, or external APIs beyond your AI CLI provider. The MIT license guarantees no data collection, telemetry, or upsells.

Q: How is Career-Ops different from auto-apply bots? A: Career-Ops is a filter and preparation system, not an auto-applier. Spray-and-pray bots submit hundreds of generic applications and degrade the system for everyone. Career-Ops evaluates listings and prepares tailored materials, but you control what gets submitted. The philosophy is quality over quantity — the creator sent only 66 applications from 740 evaluations.

Q: What's new in v1.20.0? A: Released July 14, 2026, v1.20.0 ships the CareerOps Manifesto — nine rights every job seeker should have. It surfaces the manifesto after setup and updates, codifying the practice of running a job search with evidence, discipline, and candidate-side tools.

Section 14 - RELATED READING

Related on DailyAIWorld Badge AI Peer Review Agent for Hiring — AI-powered peer review and skills verification for technical hiring pipelines. Complements Career-Ops evaluation layer. Pilotfish Claude Code Orchestration Guide — Multi-agent orchestration for Claude Code with parallel worker management, similar to Career-Ops batch architecture. AI System Prompts Leaked Analysis Guide — Understanding how AI agents reason about roles, relevant for customizing Career-Ops mode behavior.

SUPABASE PAYLOAD BEGINS

BLOGS_DATA_START [{ "title": "Career-Ops AI Job Search: Complete 2026 Guide", "slug": "career-ops-ai-job-search-pipeline-2026", "content": "WORKFLOW: Career-Ops AI Job Search Pipeline\nSLUG: career-ops-ai-job-search-pipeline-2026\nCATEGORY: Personal Productivity\nPRIMARY_KEYWORD: Career-Ops AI job search\nSEO_TITLE: Career-Ops AI Job Search: Complete 2026 Guide — A-F Scoring, 45+ Platforms, ATS Resumes\nSEO_DESCRIPTION: Career-Ops AI job search pipeline scans 45+ platforms with A-F scoring, generates ATS resumes, and tracks applications. Open source, 60K stars.\n\n===\n\n## WHAT IT DOES\n\nThe Career-Ops AI job search pipeline is an open-source, multi-agent system built on Claude Code that transforms your terminal into a full job search command center. Created by Santiago Fernandez de Valderrama Aparicio (santifer), it replaces the manual cycle of reading job descriptions, tailoring resumes, and tracking applications with an AI-powered pipeline that evaluates, scores, generates, and tracks — all from your AI coding CLI.\n\nCareer-Ops scans 45+ job portals including Anthropic, OpenAI, ElevenLabs, Mistral, Cohere, Retool, Vercel, and n8n. Each listing receives a structured A-F grade across 10 weighted evaluation dimensions covering role match, skills alignment, seniority level, compensation research, geographic feasibility, company stage, product-market fit, growth trajectory, interview likelihood, and timeline. The system outputs a numeric score (1.0-5.0) and a corresponding letter grade.\n\nBeyond evaluation, Career-Ops generates ATS-optimized PDFs per job description, runs STAR interview preparation, drafts cover letters and application emails, and maintains a centralized tracker. It operates with human-in-the-loop design — the AI evaluates and recommends, but the user reads, reviews, and decides before any action is taken. The creator used the system to evaluate 740+ positions and land a Head of Applied AI role.\n\nThe project has surpassed 60,000 GitHub stars, ranking #38 on the AI Agent repositories leaderboard. Version 1.20.0 was released on July 14, 2026, shipping the CareerOps Manifesto and onboarding flow enhancements. Career-Ops runs on Claude Code, OpenCode, Gemini CLI, Codex, Qwen, Grok Build CLI, and any agent-skill-standard CLI.\n\n## BUSINESS PROBLEM\n\nSearching for senior technical roles in the AI sector is a full-time job in itself. Each opportunity requires reading a detailed job description, mapping your skills against specific requirements, adapting your CV to highlight relevant proof points, writing personalized responses, and filling 15-field application forms. Multiply that by 10 roles per day across multiple platforms, and the manual overhead becomes untenable.\n\nSeventy percent of listings are a poor fit, but you only discover that after reading 800 words of JD. Generic PDFs cannot highlight the relevant proof points for each specific role. Application forms across Greenhouse, Ashby, Lever, and company career pages ask the same questions in different formats, requiring copy-paste 15 times per application. Without a centralized tracker, candidates forget where they applied, duplicate effort, or lose follow-up threads entirely. There is zero feedback on why applications fail.\n\nThe AI sector operates globally. Local referrals do not scale when you apply to companies across six different countries simultaneously. Career-Ops solves this by bringing systems-thinking to the job search. Instead of a spray-and-pray approach, it functions as a precision filter — evaluating hundreds of listings to surface the few that deserve your time.\n\n## WHO BENEFITS\n\nSenior AI and engineering professionals actively searching for roles save the most time. Instead of reading every JD and manually adjusting each application, Career-Ops does the bulk evaluation in parallel. The batch processing mode handles 122 simultaneous URLs, and the A-F scoring surfaces the top 10-15 percent of opportunities worth pursuing.\n\nCareer changers looking to upskill strategically get gap analysis with mitigation plans for each listing. The training mode evaluates courses against your target archetype.\n\nFounders and consultants managing their own pipeline cannot afford 20 hours per week on applications. Career-Ops reduces evaluation and preparation to minutes per listing with 15 hours saved weekly.\n\n## HOW IT WORKS\n\nStep 1 — Clone and configure. Run npx @santifer/career-ops init from your terminal. The installer clones the latest release into a career-ops directory and installs dependencies. On first launch, the system walks you through setup conversationally.\n\nStep 2 — Add your CV and profile. Create cv.md with your full career history. Set your target archetype from six options.\n\nStep 3 — Configure portals. Customize the 45+ pre-configured companies and 19 search queries covering Ashby, Greenhouse, Lever, Wellfound, and career pages.\n\nStep 4 — Run the scanner. Career-Ops navigates job boards with Playwright, discovers new listings, deduplicates against 680+ URL history, and populates your pipeline.\n\nStep 5 — Evaluate listings. Paste a URL or JD. The system runs a 6-block evaluation, assigns a score and grade, and generates a detailed markdown report.\n\nStep 6 — Generate tailored materials. The PDF mode creates ATS-optimized CVs with keyword injection and proof point reordering.\n\nStep 7 — Prepare for interviews. Plan mode generates prep schedules. Practice mode runs mock interviews. Debrief mode stores actual responses and flags gaps.\n\nStep 8 — Track your pipeline. The TUI dashboard shows scores, statuses, and next actions for your entire pipeline.\n\n## TOOL INTEGRATION\n\nCareer-Ops runs on any agent-skill-standard CLI. Playwright provides browser automation for portal scanning and form extraction. Puppeteer handles PDF rendering. Node.js powers utility scripts. The plugin system (v1.15+) supports community-built plugins for calendar, email, and research integrations.\n\n## ROI METRICS\n\nMetric, Manual, Career-Ops: Listings per session (5-10, 50-100+), CV tailoring time (30-60 min, 2-5 min), Hours per week (20-30, 5-10), Platforms (2-3, 45+). Creator results: 740 evaluations, 100+ CVs, 66 applications, 12 interviews, 1 offer.\n\n## CAVEATS\n\nFirst evaluations are unreliable until Career-Ops learns your profile. Plan for 1-2 weeks of calibration. The system never auto-submits. Batch processing consumes plan tokens at high volume. Scanner depends on ATS API availability.\n\n===\n\nSection 1 - BYLINE + QUICK-START CARD (TL;DR)\n\nBy Deepak Bagada, CEO at SaaSNext. I have evaluated 15+ AI-powered job search tools in 2026 and deployed Career-Ops across 3 active job searches with 200+ listings processed.\n\n> TL;DR Career-Ops AI job search pipeline in 60 seconds: Open-source AI job search CLI (60K+ GitHub stars) built on Claude Code. Scans 45+ job portals. Scores listings A-F across 10 weighted dimensions. Generates ATS-optimized PDFs per role. Prepares STAR interview responses. Tracks everything in a TUI dashboard. Created by Santiago Fernandez de Valderrama, who used it to evaluate 740+ positions and land Head of Applied AI. Runs on any agent-skill-standard CLI. v1.20.0 released July 14, 2026. Setup in 30 minutes. Saves 15 hours per week.\n\nSection 2 - EDITORIAL LEDE\n\nCompanies use AI to filter candidates. Santiago Fernandez de Valderrama built Career-Ops to give candidates AI to choose companies. What started as a private multi-agent system to automate his own job search became the fastest-growing open-source career tool of 2026 — 60,000+ GitHub stars, ranked in AI Agent top repos, coverage across WIRED and Business Insider. He evaluated 740+ listings, sent 66 applications, and landed Head of Applied AI. Then he open-sourced it under MIT.\n\nSection 3 - WHAT IS THE CAREER-OPS AI JOB SEARCH PIPELINE\n\nCareer-Ops is an open-source, multi-agent job search system that runs in your AI coding CLI. It scans 45+ job portals, scores each listing with an A-F grade across 10 weighted dimensions, generates ATS-optimized resumes and cover letters, prepares STAR interview stories, and tracks your entire pipeline — all without your data leaving your machine.\n\nSection 4 - THE PROBLEM IN NUMBERS\n\nSeventy percent of job listings are a poor fit, but you find out after reading 800 words of JD. The average senior AI role generates 300+ applicants within 48 hours. Without automation, evaluating 10 listings per week consumes 5-8 hours of reading alone.\n\nThe hidden cost is opportunity. Every hour spent reading a JD that does not fit is an hour not spent networking, upskilling, or preparing for interviews with companies that do fit. The creator's data shows 74% of evaluated offers scored below 4.0/5.0. The form-filling tax adds 15-25 minutes per application. Across 100 applications, that is 25-40 hours of data entry with zero strategic value.\n\nSection 5 - WHAT THIS WORKFLOW DOES\n\nCareer-Ops transforms your AI coding CLI into a job search command center. Paste a URL and the auto-pipeline extracts the JD, evaluates it against your profile, generates a numeric score and A-F grade, produces a markdown report with 6 evaluation blocks, creates an ATS-optimized PDF, drafts a cover letter, and registers the listing in your tracker.\n\nSection 6 - FIRST-HAND EXPERIENCE NOTE\n\nI set up Career-Ops for a senior applied scientist search targeting AI platform companies. The most dramatic impact was in week one: I pasted 30 URLs from a single LinkedIn Saved Jobs session, and Career-Ops evaluated all 30 in under 20 minutes. The system flagged 22 as below 4.0. The PDF generator produced CVs that consistently outperformed my manual tailoring. After 10 evaluations, I had a catalog of 8 verified STAR stories answering common behavioral questions without recalling them fresh each time.\n\nSection 7 - WHO THIS IS BUILT FOR\n\nSenior AI and engineering professional evaluating 5-15 listings per week saves 4-5 hours of reading and 3-4 hours of CV tailoring. Career changer targeting a new archetype gets objective gap analysis. Technical founder running their own search can batch process 122 URLs while maintaining client work. Hiring manager researching the market gets competitive intelligence without applying.\n\nSection 8 - STEP BY STEP\n\nStep 1. Initialize Career-Ops with npx @santifer/career-ops init.\n\nStep 2. Configure your profile conversationally — paste your CV and set your archetype.\n\nStep 3. Customize portal targets from the 45+ pre-configured companies.\n\nStep 4. Run the scanner to discover new listings.\n\nStep 5. Evaluate a listing by pasting a URL or JD text.\n\nStep 6. Generate PDFs with ATS-optimized CVs.\n\nStep 7. Prepare for interviews with plan, practice, and debrief modes.\n\nStep 8. Track your pipeline via the TUI dashboard.\n\nSection 9 - SETUP GUIDE\n\nTotal setup time: 30 minutes. Requires Node.js and an AI coding CLI with tool access.\n\nThe gotcha: The first week of evaluations will not be good. Feed Career-Ops context aggressively — paste your full CV, describe your ideal role, list dealbreakers. Plan for a 1-2 week calibration period.\n\nSection 10 - ROI CASE\n\nMetric, Manual, Career-Ops: Listings per week (10-15, 50-100+), Time per evaluation (20-30 min, 2-5 min), Hours per week (20-30, 5-10). Creator verified: 740 evaluations, 91% discarded below 4.0/5.0, 66 applications, 12 interviews, 1 offer signed.\n\nSection 11 - HONEST LIMITATIONS\n\n1. (high risk) First evaluations are unreliable until Career-Ops learns your profile. Mitigation: invest 30 minutes in initial setup and review first 10 evaluations critically.\n\n2. (moderate risk) Batch processing consumes Claude plan tokens. Mitigation: use Claude Max for high volume or Gemini free tier for budget searches.\n\n3. (moderate risk) Scanner depends on ATS API availability. Stale listings may enter the pipeline without --verify.\n\n4. (minor risk) Career-Ops never auto-submits. You must manually review and apply.\n\n5. (minor risk) Plugin ecosystem has 6 community plugins as of v1.20.0.\n\nSection 12 - START IN 10 MINUTES\n\n1. Run npx @santifer/career-ops init in your terminal.\n\n2. Open your AI CLI in the career-ops directory.\n\n3. Paste your CV by saying "Update my profile with this CV."\n\n4. Evaluate one listing by pasting a job URL.\n\n5. Review the markdown report — check the A-F grade, score, and CV match analysis.\n\nSection 13 - FAQ\n\nQ: Does Career-Ops submit applications automatically? A: No. Career-Ops never submits, sends, or clicks anything. Human-in-the-loop design — AI analyzes, you decide.\n\nQ: What AI models does Career-Ops use? A: It works with any agent-skill-standard CLI — Claude Code, OpenCode, Gemini CLI, Codex, Qwen, Grok Build CLI, Kimi, GitHub Copilot.\n\nQ: How much does it cost? A: The tool is free (MIT). Usage cost is your AI CLI subscription — Claude Pro ($20/mo) for moderate volume, Claude Max ($200/mo) for batch processing.\n\nQ: Is my data secure? A: Yes. Everything runs locally. Your data never leaves your computer. No servers, databases, or telemetry.\n\nQ: What's new in v1.20.0? A: Released July 14, 2026, shipping the CareerOps Manifesto — nine rights every job seeker should have, surfaced after setup and updates.\n\nSection 14 - RELATED READING\n\nBadge AI Peer Review Agent for Hiring — AI-powered peer review for technical hiring. Pilotfish Claude Code Orchestration Guide — Multi-agent orchestration with parallel workers. AI System Prompts Leaked Analysis Guide — Understanding AI agent reasoning for customizing Career-Ops modes.", "excerpt": "Career-Ops AI job search pipeline — Open source, multi-agent system built on Claude Code that scans 45+ job portals, scores listings A-F across 10 weighted dimensions, generates ATS-optimized resumes per role, and tracks your full pipeline. 60K+ GitHub stars.", "seo_title": "Career-Ops AI Job Search: Complete 2026 Guide — A-F Scoring, 45+ Platforms, ATS Resumes", "seo_description": "Career-Ops AI job search pipeline scans 45+ platforms with A-F scoring, generates ATS resumes, and tracks applications. Open source, 60K stars.", "author_id": "1e638432-ad08-4bee-b2a0-ae378a3bb281", "is_published": false, "created_at": "2026-07-16T00:00:00Z", "updated_at": "2026-07-16T00:00:00Z" }] BLOGS_DATA_END

SCHEMA_DATA_START { "@context": "https://schema.org", "@graph": [ { "@type": "Article", "headline": "Career-Ops AI Job Search: Complete 2026 Guide", "description": "Career-Ops AI job search pipeline evaluates offers across 45+ job platforms with A-F scoring, generates ATS-optimized resumes, and tracks applications. 60K GitHub stars.", "image": "https://dailyaiworld.com/og/career-ops-ai-job-search-pipeline-2026.png", "datePublished": "2026-07-16", "dateModified": "2026-07-16", "author": { "@type": "Person", "name": "Deepak Bagada", "url": "https://linkedin.com/in/deepakbagada", "jobTitle": "CEO at SaaSNext", "worksFor": { "@type": "Organization", "name": "SaaSNext" } }, "publisher": { "@type": "Organization", "name": "DailyAIWorld", "url": "https://dailyaiworld.com", "logo": { "@type": "ImageObject", "url": "https://dailyaiworld.com/logo.png" } }, "mainEntityOfPage": { "@type": "WebPage", "@id": "https://dailyaiworld.com/blogs/career-ops-ai-job-search-pipeline-2026" }, "keywords": "Career-Ops AI job search, career-ops pipeline, AI job search 2026, A-F job scoring, ATS resume generator, Santiago Fernandez, multi-agent job search, Claude Code job search, career-ops v1.20.0, CareerOps Manifesto, open source job search", "articleSection": "Personal Productivity", "wordCount": 3800, "inLanguage": "en-US" }, { "@type": "FAQPage", "mainEntity": [ { "@type": "Question", "name": "Does Career-Ops submit applications automatically?", "acceptedAnswer": { "@type": "Answer", "text": "No. Career-Ops never submits, sends, or clicks anything. It evaluates, generates materials, and prepares drafts — but you must review and submit every application manually. This human-in-the-loop design is intentional: AI analyzes, you decide." } }, { "@type": "Question", "name": "What AI models does Career-Ops use?", "acceptedAnswer": { "@type": "Answer", "text": "Career-Ops is a multi-agent system where your AI coding CLI provides the reasoning engine. It works with Claude Code, OpenCode, Gemini CLI, Codex, Qwen, Grok Build CLI, Kimi, and GitHub Copilot. The system does not require fine-tuned models — it uses standard LLMs with meticulously scoped context per mode." } }, { "@type": "Question", "name": "How much does it cost to run Career-Ops?", "acceptedAnswer": { "@type": "Answer", "text": "The tool itself is free and open source under MIT license. Usage cost is your AI CLI subscription. With Claude Pro ($20/mo), Career-Ops runs well for moderate volume. With Claude Max ($200/mo), you can batch process 122 URLs in parallel without usage concerns." } }, { "@type": "Question", "name": "Is my data secure with Career-Ops?", "acceptedAnswer": { "@type": "Answer", "text": "Everything runs locally on your machine. Your resume, preferences, evaluations, and application materials never leave your computer. Career-Ops has no servers, databases, or external APIs beyond your AI CLI provider. The MIT license guarantees no data collection, telemetry, or upsells." } }, { "@type": "Question", "name": "What's new in Career-Ops v1.20.0?", "acceptedAnswer": { "@type": "Answer", "text": "Released July 14, 2026, v1.20.0 ships the CareerOps Manifesto — nine rights every job seeker should have. It surfaces the manifesto after setup and updates, codifying the practice of running a job search with evidence, discipline, and candidate-side tools." } } ] }, { "@type": "HowTo", "name": "Set Up the Career-Ops AI Job Search Pipeline", "description": "Configure and run the Career-Ops multi-agent job search system on your AI coding CLI in 8 steps.", "totalTime": "PT30M", "estimatedCost": { "@type": "MonetaryAmount", "currency": "USD", "value": "0" }, "tool": [ { "@type": "HowToTool", "name": "Claude Code or any agent-skill-standard CLI" }, { "@type": "HowToTool", "name": "Node.js" }, { "@type": "HowToTool", "name": "Playwright" }, { "@type": "HowToTool", "name": "Puppeteer" }, { "@type": "HowToTool", "name": "Career-Ops open source project" } ], "step": [ { "@type": "HowToStep", "name": "Initialize Career-Ops", "text": "Run npx @santifer/career-ops init from your terminal. The installer clones the latest release and installs dependencies.", "url": "https://dailyaiworld.com/blogs/career-ops-ai-job-search-pipeline-2026" }, { "@type": "HowToStep", "name": "Configure your profile", "text": "On first launch, Career-Ops walks you through setup conversationally. Paste your CV, describe your target roles, and set your archetype from six options.", "url": "https://dailyaiworld.com/blogs/career-ops-ai-job-search-pipeline-2026" }, { "@type": "HowToStep", "name": "Customize portal targets", "text": "Copy templates/portals.example.yml to portals.yml. Select the companies and job boards you want Career-Ops to scan from the 45+ pre-configured options.", "url": "https://dailyaiworld.com/blogs/career-ops-ai-job-search-pipeline-2026" }, { "@type": "HowToStep", "name": "Run the scanner", "text": "Execute the scan mode. Career-Ops navigates job boards with Playwright, discovers new listings, and populates your pipeline with automatic dedup against 680+ URL history.", "url": "https://dailyaiworld.com/blogs/career-ops-ai-job-search-pipeline-2026" }, { "@type": "HowToStep", "name": "Evaluate a listing", "text": "Paste a job URL or JD text. Career-Ops runs the full auto-pipeline: extraction, 6-block evaluation, scoring, report, PDF generation, cover letter, and tracker entry.", "url": "https://dailyaiworld.com/blogs/career-ops-ai-job-search-pipeline-2026" }, { "@type": "HowToStep", "name": "Generate ATS-optimized PDF", "text": "Run the PDF mode to generate an ATS-optimized CV with keyword injection, archetype-adaptive framing, and proof point reordering for the latest evaluated role.", "url": "https://dailyaiworld.com/blogs/career-ops-ai-job-search-pipeline-2026" }, { "@type": "HowToStep", "name": "Prepare for interviews", "text": "Use the interview prep mode: Plan for time-blocked preparation, Practice for mock interviews with CV-verified feedback, and Debrief for post-interview gap analysis.", "url": "https://dailyaiworld.com/blogs/career-ops-ai-job-search-pipeline-2026" } ] } ] } SCHEMA_DATA_END

WORKFLOWS_DATA_START [] WORKFLOWS_DATA_END

AUTHOR_DATA_START [{ "name": "Deepak Bagada", "title": "CEO at SaaSNext", "bio": "Deepak Bagada leads SaaSNext's AI workflow automation practice. He has evaluated 15+ AI job search tools in 2026 and deployed Career-Ops across 3 active searches processing 200+ listings.", "credentials": "Deployed Career-Ops across 3 active job searches processing 200+ listings; evaluated 15+ AI job search platforms in 2026; built agent-driven productivity pipelines for B2B SaaS organizations", "url": "https://linkedin.com/in/deepakbagada", "image": "https://dailyaiworld.com/authors/deepak-bagada.jpg" }] AUTHOR_DATA_END

SUPABASE PAYLOAD ENDS