Skip to content

data-model.md

Latest commit

 

History

History
138 lines (118 loc) · 5.43 KB

data-model.md

File metadata and controls

138 lines (118 loc) · 5.43 KB
Raw

HeadHunter Data Model (source of truth)

Architecture overview (storage layout, pipelines, MCP vs scripts): docs/engineering/ARCHITECTURE.md

There is no SQL or hosted database — the “database” is a set of JSON files under the data directory.

All entities are persisted as JSON under ${CLAUDE_PLUGIN_ROOT}/data/ (or HEADHUNTER_DATA_DIR):

  • applications.json — array of JobApplication
  • interviews.json — array of InterviewRound
  • tasks.json — array of Task
  • contacts.json — array of Contact
  • notes.json — array of Note

Every record has a generated id (string, e.g. app_<timestamp>_<rand>), plus created_date and updated_date (ISO 8601). Any write MUST bump updated_date.

Storage backend preference

  1. Base44 MCP / API — if VITE_BASE44_APP_ID + token are set, use the same entities as the web app.
  2. Local JSON store (default) — ${CLAUDE_PLUGIN_ROOT}/data/*.json.
  3. Notion databases — when Notion MCP connected (see integrations).

JobApplication

Field Type Required Notes
id string auto
company string yes
role string yes Job title
job_url string Original posting URL
status enum Default Saved. Saved, Applied, Phone Screen, Technical, Onsite, Offer, Accepted, Declined, Rejected, Ghosted
priority enum High, Medium, Low (default Medium)
applied_date date ISO date
source string LinkedIn, Referral, Company Site, etc.
salary_min number
salary_max number
currency string Default USD
location string
remote_type enum Remote, Hybrid, On-site
job_description string Markdown OK
requirements string Markdown OK
color_label string Hex for Kanban accent
notion_page_id string Set after Notion sync
research_dir string Path to data/research/<slug>/ (e.g. nvidia-senior-ai-llm-solutions)
last_research_at datetime When the interview-research pipeline last ran
match_score number 0–100, set by /headhunter:scan
success_score number 0–100, set by /headhunter:scan
scanner_notes string Verdict + key gaps from scanner
tailored_cv_path string Path to HTML tailored CV (set by /headhunter:apply)
cover_letter_path string Path to cover letter markdown
referral_contact_id string FK → Contact who referred this application
research_status enum not_started / in_progress / complete / stale. Auto-set by post-research-hook
created_date datetime auto
updated_date datetime auto bump on every write

InterviewRound

Field Type Required Notes
id string auto
application_id string yes FK → JobApplication
round_number number Auto-increment per app
round_type enum yes Recruiter Call, Phone Screen, Technical, System Design, Behavioral, Take-Home, Onsite, Panel, Final, Reference Check
scheduled_at datetime
duration_minutes number Default 60
interviewer_name string
interviewer_linkedin string
google_calendar_event_id string
google_meet_link string
status enum Scheduled, Completed, Cancelled, No-show
outcome enum Pending, Passed, Failed, Cancelled
prep_notes string
reflection_notes string
difficulty_rating number 1–5
vibe_rating number 1–5
prep_checklist array { id, text, done }[]
questions_asked string
my_answers string
what_went_well string
what_went_poorly string
overall_feeling enum Great, Good, Neutral, Nervous, Poor
key_takeaways string
mock_sessions array [{ date, verdict, avg_score, strengths[], improvements[] }] — appended by /headhunter:mock

Task

Field Type Notes
id string auto
application_id string Optional FK
title string Required
description string
due_date datetime
priority enum High, Medium, Low
completed boolean Default false
google_task_id string Sync ID
todoist_task_id string Sync ID
notion_task_page_id string Sync ID
reminder_sent boolean

Contact

Field Type Required
id string auto
application_id string yes
name string yes
role_at_company string Recruiter, Hiring Manager, etc.
email string
phone string
linkedin_url string
notes string

Note

Field Type Required
id string auto
application_id string yes
interview_round_id string optional
content string yes
note_type enum General, Follow-Up, Offer Details, Rejection Reason, Research

Event (auto-logged, powers the Timeline)

Written automatically by crud.js on create/move/complete — not edited by hand.

Field Type Notes
id string auto
ts datetime when it happened
type string application_created, status_change, interview_added, task_added, task_completed, note_added
application_id string FK → JobApplication
summary string human-readable line
meta object type-specific (e.g. {from, to} for status_change)