santifer/career-ops
career-ops shows how a coding-agent harness can run personal operations: structured scoring, parallel workers, document generation, and human approval around a durable data store.
career-ops turns coding CLIs into a job-search pipeline with **14 skill modes**, structured listing scores, ATS-ready PDFs, portal scanning, and headless batch workers. The author reports evaluating **740+ listings** and producing **100+ tailored CVs**.
The reusable pattern is broader than recruiting: keep agent modes in shared files, use one tracked source of truth, delegate batches to subagents, and place explicit review gates before consequential actions. Codex can invoke modes through plain-language prompts or headless exec calls.
career-ops turns coding CLIs into a job-search pipeline with **14 skill modes**, structured listing scores, ATS-ready PDFs, portal scanning, and headless batch workers. The author reports evaluating **740+ listings** and producing **100+ tailored CVs**. The reusable pattern is broader than recruiting: keep agent modes in shared files, use one tracked source of truth, delegate batches to subagents, and place explicit review gates before consequential actions. Codex can invoke modes through plain-language prompts or headless exec calls. Early evaluations depend heavily on the profile, CV, preferences, and proof points supplied. The system never submits applications itself, and stale ATS feeds can still introduce expired listings unless verification is enabled.
career-ops is a concrete application of coding-agent harness patterns outside software delivery: shared skills, structured state, parallel bounded work, and review gates coordinate a consequential personal workflow. It supports delegating batch analysis without delegating final action, while narrowing the evidence to author-reported volume and profile-dependent outputs rather than demonstrated accuracy or hiring outcomes.