8000 GitHub - avabbbb/OfferU: Local-first AI Career OS — research jobs, tailor resumes, track applications, and learn from every interview. · GitHub
[go: up one dir, main page]

Skip to content
avabbbbPublic

About

Local-first AI Career OS — research jobs, tailor resumes, track applications, and learn from every interview.

Topics

Resources

Contributing

Security policy

Stars

23 stars

Watchers

0 watching

Forks

Repository files navigation

OfferU logo

OfferU

Give OfferU a job. Get an evidence-backed application workspace.
See what the role really asks for, what you can prove, and what to prepare next.

A local-first AI Career OS that compounds what you learn across jobs and interviews.

English · Videos · 简体中文 · Quickstart · PRD · Docs

Local-first · Evidence-driven · Human-controlled

OfferU demo: Today → Job → Role Intelligence → Resume → Pipeline

Watch the videos

Play the 60-second OfferU launch film Play the OfferU product walkthrough
Launch film · 1 min
Why every application should start from evidence, not from zero
Product walkthrough · 2 min 40 s
Save a job → Role Intelligence → tailored resume → Pipeline

Click a cover to play the 1080p video.

Internal Beta — source available, signed installer not yet released. See Known issues.

OfferU workbench Job detail with role intelligence
Today: jobs, evidence and next actions Job detail: research, gaps, preparation
Application pipeline Resume workspace
Pipeline: stages, timeline, next action Resume: tailoring with evidence check

Why OfferU?

Most job-search tools solve one step.

Your JD lives in the browser. Your resume lives in a file. Research lives in another tab. Application tracking becomes a spreadsheet. Interview learning disappears into another AI chat.

So every application starts from zero.

OfferU changes the unit of work from a chat to a durable Job Workspace:

Target Job
    ↓
What does this role really care about?
    ×
What can I actually prove?
    ↓
What should I prepare next?
    ↓
Evidence-backed Job Workspace

That Workspace then compounds into the larger Career OS:

Career Profile
      ↓
   Save Job
      ↓
Role Intelligence
      ↓
Evidence Gap
      ↓
Tailored Resume
      ↓
Application Pipeline
      ↓
Targeted Interview
      ↓
Debrief & Learning
      ↺

Instead of opening a fresh AI chat for every job, OfferU keeps one persistent, evidence-back 8000 ed career context and carries what it learns across the entire search.

One Job, one durable Workspace

A Job is not just a detail page. It is the canonical workspace for one opportunity:

Job Snapshot
Role Intelligence
Evidence Map
Application Materials
Interview
Timeline / Next Action

Agent work progressively materializes into these objects, so closing a chat does not make the work disappear.

Persistent career context

OfferU maintains structured career evidence — experience, achievements, skills, preferences, goals and reviewed learning observations.

AI suggestions never silently become career facts. New information enters a reviewable candidate flow before it can update your long-term profile.

Role Intelligence

OfferU does not only summarize a job description. It compares the target role against a cohort of similar jobs, separates what is common from what is distinctive, and maps those signals against your own evidence.

What does this role emphasize?
           ×
What can I actually prove?
           ↓
What should I prepare next?

Evidence-grounded resume tailoring

Every target job can have its own tailored resume without overwriting the source resume.

The Resume Workspace supports structured manual editing, live A4 / Letter preview, job-specific resume versions, AI proposals with before / after diffs, accept / reject review, stale-proposal protection and PDF export.

AI-generated claims are checked against career evidence before they become trusted application content.

Application pipeline

Today, Pipeline, Job Detail and Timeline all read from the same underlying career state. Application progress is modeled as events rather than independent UI state, so OfferU can project one truth across the product instead of asking you to maintain several trackers.

Targeted interview practice

Interview preparation is grounded in the intersection of role delta, career evidence gap and previous interview learning:

Role Delta
×
Career Evidence Gap
×
Previous Interview Learning

OfferU generates targeted focus areas, runs turn-based practice, challenges vague answers, produces transcript-backed debriefs and turns useful observations into reviewable learning candidates.

Your AI, with a controlled Career OS

OfferU prefers the local AI agent you already use and allows a built-in fallback, but no model owns business truth.

External local Agent (preferred) / OfferU fallback
    ↓
OfferU Skill + optional Career Skills
    ↓
Agent Tool Surface
    ↓
Operation Registry
    ↓
Proposal / Approval
    ↓
Career Runtime

The active agent reasons. OfferU controls capabilities, evidence, side effects and persistent truth. Third-party resume/interview/recruiting Skills may contribute methodology, while sensitive changes and irreversible actions remain under OfferU and user control.


Product surfaces

Surface Purpose
Today What changed, what OfferU finished, what needs your attention, and what matters next
Pipeline Every opportunity, application stage, timeline and next action
Job Durable Job Workspace: snapshot, Role Intelligence, Evidence Map, materials, interview and timeline
Profile Long-term career evidence, goals, preferences and reviewed learning

Memory is a mechanism for evolving Profile — not a separate product silo. The Agent is a system-wide capability — not another disconnected chat window. The extension/ browser add-on (WXT) is the job-capture surface — it imports listings into the same Operation Registry instead of scraping or auto-applying on its own.


AI setup

The product has two front doors and one Career Truth.

Normal users: App-first

Install OfferU
→ auto-detect local AI
→ project/register OfferU Skill where supported
→ save a Job
→ work appears in the canonical Job Workspace

Power users: Skill-first

Open Codex / Claude Code / WorkBuddy / OpenCode / OMP / another supported host
→ enable OfferU Skill
→ "analyze this job for me"
→ Agent creates/resolves the canonical Job
→ open the same Job Workspace in OfferU

The product direction remains Connect → Auto → Ready.

Skill is the Agent entry; Job Workspace is the user product.

Normal users should not have to understand runtimes, protocol versions, model IDs or custom endpoints. OfferU detects a local agent you already have, checks it, and consumes its own model and account:

        OfferU

   AI Connection
        ↓
    Auto Detect
        ↓
┌────────────┼────────────┐
Codex  WorkBuddy  Claude  OMP / others
        ↓
   OfferU Skill
   + compatible Career Skills
        ↓
   OfferU Bridge
        ↓
 Operation Registry
        ↓
   Career Runtime

If you already use Codex with your ChatGPT account, Claude Code with your Claude account, or OpenCode, OfferU does not need an API key from you at all — the agent brings its own model and authentication.

API configuration is the fallback for users with no local agent, self-hosting users, and advanced users who deliberately want to configure the stack. It lives behind Advanced, where you get two protocols rather than dozens of vendor presets:

  • OpenAI-compatible endpoint
  • Anthropic-compatible endpoint

Runtime diagnostics, experimental providers and provider health are advanced / developer surfaces.


Architecture

One set of rules, three places it runs:

flowchart LR
    UI[React UI] -->|fetch /api/*| APP[Hono app<br/>backend-ts createApp]
    EXT[Browser extension] --> APP
    AG[Codex / Claude Code] -->|MCP stdio / CLI| APP
    APP --> CORE[packages/core<br/>tools, rules, undo, Agent loop]
    CORE --> IDB[(IndexedDB<br/>web version)]
    CORE --> SQL[(SQLite<br/>desktop companion)]
Loading
  • packages/core — pure TypeScript domain rules: the tool registry, validation, overview, change log and undo, resume rendering, the built-in Agent loop. No React, DOM or Node APIs.
  • backend-ts — one runtime-agnostic Hono app plus adapters. It runs inside the page for the web version (data in IndexedDB), as the Node desktop companion (SQLite, local files, PDF, BOSS login), and its Cloudflare Worker entry serves the static site and relays model calls. The Worker stores nothing: no D1, KV, Durable Objects or R2.
  • backend/ — the previous Python (FastAPI) backend, kept as a rollback path and for the domains not migrated yet (mail sync, interviews).

Capabilities per runtime, where code lives and how to add a tool: docs/ARCHITECTURE.md. Data model and tool semantics: docs/DESIGN.md.


Current technology

React / TypeScript   → product UI (Vite)
TypeScript + Hono    → backend: packages/core rules, backend-ts adapters
IndexedDB            → web version data (in the user's browser)
SQLite (node:sqlite) → desktop companion data
Cloudflare Workers   → static hosting + stateless model relay (free plan)
Tauri / Rust         → desktop shell
Python / FastAPI     → legacy backend, rollback path

OfferU intentionally does not duplicate business logic across UI, CLI, plugins and agent integrations.


Safety principles

  • AI output is not automatically career truth.
  • Important mutations are reviewable and auditable.
  • External irreversible actions require explicit user control.
  • Browser automation may assist with forms but must not silently submit applications.
  • Career evidence preserves provenance; behaviour signals and model inferences enter a review inbox first.
  • Provider failures must be visible rather than silently returning fake success.
  • Your data stays with you: in your browser (web version) or on your machine (desktop). The Cloudflare Worker only serves static files and relays model calls; it keeps no user data and no logs of them.
  • In the web version API keys are encrypted with a non-extractable per-browser key before they are stored in IndexedDB. The legacy Python backend keeps them in the OS keyring.
  • Credentials should stay out of model context, logs and version control.

See SECURITY.md for the current security status.


Getting started

OfferU is not yet published as a signed consumer installer.

Web version (data stays in your browser)

npm --prefix backend-ts install && npm --prefix frontend install
cd backend-ts
npm run build:web        # frontend in local-web mode
npx wrangler deploy      # free Cloudflare Workers plan; or `npm run worker:dev` locally

Open the site, add your model API key under Settings → Model, and start. See docs/DEPLOY-CLOUDFLARE.md.

Source development — docs/DEVELOPMENT.md (npm commands, tests, contract tests).


Release status

cd backend-ts && npm test && npm run typecheck
cd frontend && npm test && npm run typecheck && npm run build:web

These commands only validate their own scope; they do not mean "ready to release".


Roadmap

Current priorities are productization, not more top-level features:

  1. Zero-Setup onboarding — native install, local-Agent auto-discovery, Profile bootstrap, first Job, inbox, useful Today.
  2. Real external-Agent Eval — validate autonomous tool discovery and complete user journeys before further tool compression.
  3. Browser + inbox productization — one-click job capture and beginner-friendly read-only progress sync.
  4. Public desktop release — signed installer, clean-machine setup, migration, backup, restore and upgrade.
  5. Privacy, security and real-user iteration — keep trust gates hard while reducing user effort.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

  1. Fork the repository
  2. Create your feature branch (git checkout -b feature/amazing-feature)
  3. Commit your changes (git commit -m 'Add some amazing feature')
  4. Push to the branch (git push origin feature/amazing-feature)
  5. Open a Pull Request

See CONTRIBUTING.md for what the main repository accepts, including the rule on third-party integrations.


License

MIT

About

Local-first AI Career OS — research jobs, tailor resumes, track applications, and learn from every interview.

Topics

Resources

Contributing

Security policy

Stars

23 stars

Watchers

0 watching

Forks

Releases

Packages

Used by

Contributors

Languages

0