Skip to main content

Proposal 001 — Initial vision article

Status: Approved · Drafting
Series: ARTICLE_SYSTEM.md (series plan) Article 01
GitHub: #56 Write Article: Initial Vision (article-approved · article-draft)
Research: ../research/JOBJITSU_MEDIUM_ARTICLE_BLUEPRINT.md
Draft: ../drafts/001-initial-vision.md

Approved: 2026-07-23 (human)

Proposed title​

Selected: The Gentle Art of Landing the Job: Why We’re Building a Local-First Career OS

Alternates retained for reference:

  1. Your Résumé Isn’t Training Data: An Open-Source Case for On-Device Job Hunt AI
  2. Agent Prepares, You Send: Privacy Architecture for the AI Job Market

Reason for article​

Establish JobJitsu’s public narrative: why it exists, what it refuses to become, and how local-first Agent architecture differs from cloud career SaaS and spray-and-pray autopilot — while the repo is still in early foundation.

  • Documentation + architecture SSOT in place
  • Monorepo / docs site foundation
  • Medium article blueprint completed (docs/articles/research/)

Not tied to a shipping product wave completion yet — this is the series opener.

Technical topics​

  • Vision and promise triad (on-device / on-target / on your terms)
  • Problem + accessibility (no named competitors)
  • Local AI primary path; remote opt-in honesty
  • Architecture overview (host, events, AiProvider, Queue, Send)
  • Agent system overview (Agent ≠ Send; roles ≠ separate bots)
  • Desktop vision (Tauri + React)
  • Open source + community invitation

Required documentation​

  • docs/brand/PRODUCT_PHILOSOPHY.md
  • docs/product/PRODUCT_VISION.md
  • docs/product/NON_GOALS.md
  • docs/product/FEATURES.md (status honesty)
  • docs/architecture/OVERVIEW.md
  • docs/architecture/AI_ARCHITECTURE.md
  • docs/architecture/PACKAGE_BOUNDARIES.md
  • docs/adr/0001-tauri.md, 0005-ai-runtime.md, 0009-send-boundary.md
  • Blueprint: docs/articles/research/JOBJITSU_MEDIUM_ARTICLE_BLUEPRINT.md

Suggested diagrams​

  • System architecture (host / UI / local model)
  • Agent ≠ Send flow (prepare → queue → approve → send)
  • Docs-first development lifecycle

Labels / project​

  • Labels: article-approved, article-draft (remove article-needed when drafting)
  • Content Status: Drafting

On-device · On-target · On your terms — see brand guidelines.