# Cross-agent job-tailored resume pipeline request
AI Summary
Purpose:
- Preserve the user's request for a shared Claude Code and Codex pipeline that
tailors resumes to supplied job postings.
Key points:
- A supplied posting must drive evidence selection and resume ordering.
- Existing verified career material should be reassembled rather than rewritten
from unsupported memory.
- The user wants a 95/100 release threshold.
- The implementation interprets 95 as a reproducible document-quality gate,
not a hiring-outcome guarantee.
- Claude Code and Codex must follow the same source, template, reviewers, and
scoring rules.
Relevant when:
- Designing or revising the job-tailored resume workflow.
- Explaining why a resume package did or did not pass the 95-point gate.
Do not read full document unless:
- The original intent or score interpretation is in question.
Linked documents:
skills/tailor-resume-to-job/SKILL.mdai/wiki/concepts/job-tailored-resume-pipeline.mdai/assets/resume/README.md
Open Questions
- None.
Details
User request, 2026-07-22:
앞으로 이력서를 만들 때 공고를 주면 그 공고에 맞게 이력서를 잘 짜집기해서 점수가 95점 이상이게 하는 파이프라인을 클로드 코덱스 다 적용할 수 있게 해줘.
Implementation boundary:
- The pipeline may reorder and rewrite verified evidence for relevance.
- It may not fabricate a missing hard requirement to raise the score.
- A low candidate-to-posting fit must remain visible as an honest gap.