Career Content Authority
Career Content Authority
src/data/resume.json is the first canonical store for shared career-suite facts. The content authority manifest at src/data/content-authority.json defines which fields are public, which fields are private, which downstream targets consume projections, and which repo-local JSON commands are stable enough for Codex Platform or MCP wrappers to call.
This repo owns the resume content model, PDF generation, and resume/CV export behavior. The platform repo owns orchestration contracts only.
src/data/resume-format.json is the target-local projection of the platform resume format contract. It lists the output profiles this repo can validate for agents: ATS PDF, human PDF, web resume, targeted job application packets, and the planned ATS DOCX fallback.
JSON Command Surface
The first API surface is a repo-local command:
node scripts/career-content-api.mjs status
Every command emits a JSON envelope with:
okstatusrepo_idcommandblockersartifactsnext_actions
Supported commands:
status: validates the content authority and returns counts, public identity, and sync targetsprofile: returns the public profile and high-level resume inventorysync-preview --target=all: previews deterministic downstream payloads forresume,docs,hire, andlinktailor-preview --job-text="...": ranks existing skills, experience, and projects against a job posting without generating a PDF or inventing factsjob-normalize --job-text="...": normalizes a posting into deterministic job fields, keywords, and detected requirement termsfit-report --job-text="...": scores job overlap against canonical resume facts and exposes gapsselection-plan --job-text="...": creates a source-backed section plan fortargeted_job_applicationreview-packet --job-text="...": bundles normalized job data, fit report, selection plan, claim audit, and risk flagsformat-status: returns output profiles, source-fact counts, and format contract pointersformat-validate-source: validates canonical resume content and source fact IDsformat-validate-export --profile=web_resume: validates one output profile against current source and artifactsformat-claim-audit --targeted-packet-json='{"source_fact_ids":["identity.name"]}': checks selected fact IDs in a targeted packet
Publish Workflow
The current publish path remains:
- update
src/data/resume.json - run
npm run career:status - run
npm run career:sync-preview -- --target=all - run
npm run resume.format.status - run
npm run resume.format.validate-source - run
npm run resume.format.validate-export -- --profile=web_resume - run
npm run career:review-packet -- --job-text="..."before any targeted application use - run
npm run check - run
npm run build - run
npm run pdfwhen a refreshed PDF is needed
Tailored resume PDF generation should consume a reviewed review-packet after the selection plan and claim audit pass.
Application Automation Boundary
Job polling, job-fit scoring, application packet generation, and application submission are separate later workers. This slice only prepares the content and selection substrate. Automatic submission must remain deferred until there is explicit policy for identity use, claim review, destination allowlists, and human approval.