Skip to content
Back to search

job-hunter-os

Zero-Code AI Career Co-Pilot: 48 pure-Markdown skills + zero-dependency Node CLI for fact-checked CV tailoring, Recruiter-SEO audits, and human-approved ATS prep.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 19 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/redasalem/veycraft with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/redasalem/veycraft'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/redasalem--job-hunter-os/badge.svg)](/listings/redasalem--job-hunter-os)

HTML

<a href="/listings/redasalem--job-hunter-os"><img src="/api/listings/redasalem--job-hunter-os/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Works on **macOS, Linux, and Windows** with Node.js `>= 18`:

# 1. Instant offline CV hygiene audit (optional --share writes privacy-safe SVG/MD card to logs/exports/)
npx github:redasalem/job-hunter-os audit ./resume.md --share

# 2. Initialize globally (~/.job-hunter-os + IDE skill/command auto-discovery)
npx github:redasalem/job-hunter-os init --global

# 3. Open your AI IDE:
#    - Run `/demo`         (instant synthetic workflow, zero setup)
#    - Run `/quick-tailor` (your real CV + 1 JD, no browser login)

# 4. Optional (Full Career OS only): sign in to local browser-profile/ and run /setup
npx github:redasalem/job-hunter-os login --global
npx github:redasalem/job-hunter-os doctor --global

> **Zero-Dependency Core Architecture:** `package.json` maintains strictly **`"dependencies": {}`** and **`"devDependencies": {}`**. The core CLI uses Node.js built-ins only. Optional browser automation and PDF export use pinned Playwright MCP (`@playwright/mcp@0.0.86`) invoked via `npx` only when you use browser-assisted workflows; Notion sync is opt-in and off by default. Alternative OS install scripts: [`scripts/install.sh`](scripts/install.sh) and [`scripts/install.ps1`](scripts/install.ps1).

---

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 226 lines
  1. # Job Hunter OS
  2. ### Truthful CV Tailoring for the AI Agent You Already Use.
  3. Give your existing AI coding agent (**Cursor, Claude Code, Antigravity, Windsurf, Cline, Gemini CLI, Codex, OpenCode**) one CV and one job description. Get a fact-checked tailored HTML/TXT resume, a transparent evidence map, an ATS keyword comparison, and an explicit list of missing skills.
  4. No fabricated experience. No blind applications. No external API keys. You remain in control.
  5. **48 Markdown Skills** · **0 Core npm Dependencies** · **Local-First** · **Human-Controlled**
  6. **[Explore the CV Tailoring Demo ↗](https://redasalem.github.io/job-hunter-os/)** · **[Quick Installation](#installation)** · **[العربية (RTL)](README.ar.md)**
  7. [![Zero Core Dependencies](https://img.shields.io/badge/Core_CLI-0_NPM_Dependencies-2ea44f)](package.json)
  8. [![48 Markdown Skills](https://img.shields.io/badge/Skills-48_Markdown_Skills-167d60)](skills/REGISTRY.md)
  9. [![MIT License](https://img.shields.io/badge/License-MIT-2ea44f)](LICENSE)
  10. ```bash
  11. # Try an instant offline hygiene audit on your own resume right now (zero setup, local-only)
  12. npx github:redasalem/job-hunter-os audit ./resume.md --share
  13. ```
  14. ---
  15. ## Actual CV Tailoring Output (Synthetic Example)
  16. All names, employers, and metrics below come from synthetic repository fixtures ([`evals/fixtures/`](evals/fixtures/candidate-profile.md)) to illustrate the exact output structure produced by `/demo` and `/quick-tailor`:
  17. ![Job Hunter OS — Truthful CV Tailoring Output Preview (Synthetic Demonstration Data)](assets/demo-preview.png)
  18. | Input You Provide | Output You Receive |
  19. |---|---|
  20. | Existing CV (`.md` / `.txt` / pasted text) + confirmed facts | Fact-checked single-column tailored resume (**HTML + TXT**, optional PDF) |
  21. | One target job description (or batch during `/daily-run`) | Line-by-line **Evidence Map**, **ATS Keyword Comparison**, and explicit **True Gaps** |
  22. | Optional local browser session (only for full Career OS) | **Interview Dossier**, `<= 75`-word draft outreach, Live DOM Read-Back Table before `APPROVE` |
  23. ---
  24. ## Before vs. After Transformation
  25. ```diff
  26. - Built backend APIs for an e-commerce and payments app using Node.js.
  27. + Built idempotent Node.js and TypeScript webhook processors backed by PostgreSQL and AWS SQS,
  28. + reducing payment webhook p95 latency by 38% across $4.2M daily settlement volume [verified: candidate-profile.md].
  29. ! Blocked claim: "Managed multi-cluster Kubernetes" (0 verified years — isolated under True Gaps)
  30. [ATS Keyword Coverage: 71.4% -> 92.8% (heuristic comparison, not an ATS guarantee)]
  31. [Status: dry_run: true · Awaiting explicit job-specific human APPROVE]
  32. ```
  33. - **Zero invented claims:** Employers, titles, dates, tools, and metrics must trace back to your verified facts. Missing skills are never quietly added—they are listed under **True Gaps**.
  34. - **Anti-AI slop filter:** Strips generic clichés (`delve`, `spearheaded`, `synergy`, `testament`) and prose em-dashes in favor of plain engineering language.
  35. - **No blind submissions:** Every live ATS form submission requires an explicit, exact, job-specific human `APPROVE` after inspecting the **Live DOM Read-Back Table**. No score-based auto-submit, no batch approvals across jobs.
  36. ---
  37. ## What Makes the Workflow Trustworthy
  38. Job Hunter OS is transparent about **which layer enforces each control** (details in [`docs/execution-safety.md`](docs/execution-safety.md)):
  39. | Control Layer | What It Enforces | Concrete Examples |
  40. |---|---|---|
  41. | **Deterministic CLI Enforcement (`bin/cli.mjs`)** | Verified locally by zero-dependency Node.js code with exit codes | `pre-submit` (`<= 75`-word outreach cap & daily/Days 1–3 warm-up caps), `.git/hooks/pre-commit` PII & token blocker, JSON integrity in `backup`, live-PID check in `unlock`, SSRF guard in `doctor`/`audit`, XSS-escaped JSON in `dashboard`, SSH commit signature verification in `update`, and static fixture grading in `eval`/`proof` |
  42. | **Agent Workflow Requirements (`AGENTS.md`, `skills/*.md`)** | Structured instructions executed by your AI model inside your IDE *(not an independent OS sandbox)* | Fact-checking CV claims against verified profile, isolating missing skills in **True Gaps**, treating scraped JDs as `UNTRUSTED_EXTERNAL_DATA`, upload-first ATS filling, Live DOM Read-Back verification, and unconditional per-job `APPROVE` pause |
  43. > **Privacy & Automation Honesty:** CVs, logs, generated resumes, and optional `browser-profile/` stay local on your machine (zero telemetry, zero SaaS backend). Your chosen AI IDE/model processes chat prompts under its own provider policy—use a local model if strict offline privacy is required. Browser automation never bypasses CAPTCHAs, 2FA, or rate limits; any challenge trips a 24-hour platform cooldown and stops immediately.
  44. ---
  45. ## Fastest Ways to Get Started
  46. You do **not** need browser login or full platform setup to see value:
  47. | Step | Slash Command | Purpose | Personal CV / Browser Login? |
  48. |---|---|---|---|
  49. | **1. Instant Sandbox** | `/demo` | Run a 15-second synthetic simulation across 3 sample JDs ([`examples/sandbox-demo/`](examples/sandbox-demo/sample-jds.md)) with an immediate tailored preview for Job #1 | **None required** |
  50. | **2. First Real Tailored CV** | `/quick-tailor` | Paste your real CV + 1 JD, confirm extracted facts, and get a tailored HTML/TXT CV, Evidence Map, ATS comparison, and True Gaps | **CV + 1 JD only** (no browser login) |
  51. | **3. Full Career OS (Optional)** | `login` → `/setup` → `/daily-run` | Multi-platform job discovery, ghost-job filtering, interview dossiers, and human-approved ATS form preparation | **Browser login + 5 readiness gates** |
  52. ---
  53. ## Installation
  54. Works on **macOS, Linux, and Windows** with Node.js `>= 18`:
  55. ```bash
  56. # 1. Instant offline CV hygiene audit (optional --share writes privacy-safe SVG/MD card to logs/exports/)
  57. npx github:redasalem/job-hunter-os audit ./resume.md --share
  58. # 2. Initialize globally (~/.job-hunter-os + IDE skill/command auto-discovery)
  59. npx github:redasalem/job-hunter-os init --global
  60. # 3. Open your AI IDE:
  61. # - Run `/demo` (instant synthetic workflow, zero setup)
  62. # - Run `/quick-tailor` (your real CV + 1 JD, no browser login)
  63. # 4. Optional (Full Career OS only): sign in to local browser-profile/ and run /setup
  64. npx github:redasalem/job-hunter-os login --global
  65. npx github:redasalem/job-hunter-os doctor --global
  66. ```
  67. > **Zero-Dependency Core Architecture:** `package.json` maintains strictly **`"dependencies": {}`** and **`"devDependencies": {}`**. The core CLI uses Node.js built-ins only. Optional browser automation and PDF export use pinned Playwright MCP (`@playwright/mcp@0.0.86`) invoked via `npx` only when you use browser-assisted workflows; Notion sync is opt-in and off by default. Alternative OS install scripts: [`scripts/install.sh`](scripts/install.sh) and [`scripts/install.ps1`](scripts/install.ps1).
  68. ---
  69. ## Product Architecture
  70. ```text
  71. JOB HUNTER OS
  72. |
  73. +----------------+----------------+
  74. | | |
  75. Skills CLI Integrations
  76. | | |
  77. 48 Markdown Node.js (0 deps) AI IDEs / Optional MCP
  78. (Progressive audit / proof / Cursor, Claude Code,
  79. Disclosure) eval / pre-submit Antigravity, Codex, etc.
  80. |
  81. CV + Job Description
  82. |
  83. Human Fact Confirmation & Verification
  84. |
  85. ATS Keyword Comparison + Evidence Map + True Gaps
  86. |
  87. Human Review (HTML / TXT / Optional PDF)
  88. |
  89. [Optional Browser Prep] Live DOM Read-Back Table
  90. |
  91. Explicit Job-Specific Human APPROVE (dry_run: true default)
  92. ```
  93. - **Markdown-first runtime:** Your existing AI IDE/model reads `AGENTS.md` and `skills/REGISTRY.md`, loading only the exact skill needed for the active step ([`docs/ARCHITECTURE.md`](docs/ARCHITECTURE.md)).
  94. - **Zero-dependency CLI:** Handles cross-OS scaffolding, IDE skill syncing, deterministic pre-submit checks, offline CV hygiene audits, local dashboard generation, backups, and fixture evaluation (`bin/cli.mjs`).
  95. ---
  96. ## Technical Proof & Evaluation Scope
  97. Regenerate the proof card locally straight from `package.json`, `skills/skills-manifest.json`, and the deterministic fixture validator:
  98. ```bash
  99. npx github:redasalem/job-hunter-os proof
  100. ```
  101. ![Job Hunter OS Reproducible Proof Card](assets/proof-card.svg)
  102. [![CI Fixture Validator 20/20](https://img.shields.io/badge/CI_Fixture_Validator-20%2F20_Baseline-2ea44f)](evals/README.md)
  103. [![Optional Playwright MCP](https://img.shields.io/badge/Optional_Browser-Playwright_MCP-167d60)](mcp/README.md)
  104. [![agentskills.io Standard](https://img.shields.io/badge/Standard-agentskills.io-2ea44f)](.agents/skills/daily-run/SKILL.md)
  105. > **Honest Eval Scope:** `eval --baseline` and `proof` perform **deterministic fixture & schema validation** against immutable [`evals/fixtures/ground-truth.json`](evals/fixtures/ground-truth.json)—they are **not** an independent live-model accuracy benchmark. External model runs are stored separately (e.g., [`evals/results/live-agent-eval-run.json`](evals/results/live-agent-eval-run.json)) and graded with `eval --file <json>`. Reproduce on any model with [`evals/PROMPT.md`](evals/PROMPT.md).
  106. <details>
  107. <summary><strong>Inspect all 20 deterministic safety & workflow assertions (EVAL-01..EVAL-20)</strong></summary>
  108. | Validator Group | Fixtures Checked | Verified Invariants | Status |
  109. |---|---|---|---|
  110. | **Anti-Hallucination & ATS Tailoring** (`EVAL-01, 04..07`) | [`candidate-profile.md`](evals/fixtures/candidate-profile.md), [`ground-truth.json`](evals/fixtures/ground-truth.json), [`01-strong-match.md`](evals/fixtures/jds/01-strong-match.md) | `0` invented metrics, `Kubernetes` isolated in `true_gaps`, `0` banned AI clichés/em-dashes, 1-page A4 budget (`<= 38` lines) | **5 / 5 PASS** |
  111. | **Prompt-Injection & Honeypot Defense** (`EVAL-03, 08, 17`) | [`02-prompt-injection-trap.md`](evals/fixtures/jds/02-prompt-injection-trap.md), [`01-strong-match.md`](evals/fixtures/jds/01-strong-match.md) | Quarantines hidden `SYSTEM OVERRIDE` (`0` leaked files, salary floor untampered), leaves `website_url_honeypot` blank, captures `PINEAPPLE` codeword | **3 / 3 PASS** |
  112. | **Ghost, Blacklist & Cross-Channel Dedup** (`EVAL-02, 09..12`) | [`03-ghost-expired-job.md`](evals/fixtures/jds/03-ghost-expired-job.md), [`04-blacklist-duplicate.md`](evals/fixtures/jds/04-blacklist-duplicate.md), [`05-stretch-mismatch.md`](evals/fixtures/jds/05-stretch-mismatch.md) | Skips `>90d` closed jobs, blocks `blacklisted_companies`/`blacklisted_domains`, deduplicates `acme-cloud::senior-backend-engineer` | **5 / 5 PASS** |
  113. | **ATS Form Mocks, DOM Read-Back & Anti-Ban** (`EVAL-13..16, 18..20`) | [`greenhouse-mock.html`](evals/fixtures/forms/greenhouse-mock.html), [`lever-mock.html`](evals/fixtures/forms/lever-mock.html), [`ashby-mock.html`](evals/fixtures/forms/ashby-mock.html) | Upload-First (`resume.pdf`), neutral EEO defaults, `gdpr_consent: true`, Live DOM Read-Back Table before `APPROVE`, and 24h circuit-breaker cooldown | **7 / 7 PASS** |
  114. </details>
  115. ### ATS & Platform Verification Matrix
  116. | Support Tier | Platforms / ATS Engines | Verification Evidence |
  117. |---|---|---|
  118. | **Fixture-Verified** | **Greenhouse**, **Lever**, **Ashby** | Self-contained HTML form fixtures ([`evals/fixtures/forms/`](evals/fixtures/forms/greenhouse-mock.html)) covering file upload, honeypots, EEO/GDPR & DOM read-back |
  119. | **Playbook-Ready / Field-Testing** | **Workday** *(manual fallback)*, **SmartRecruiters**, **BreezyHR**, **LinkedIn Jobs**, **Wellfound** | Markdown playbooks with `<60d` `last_verified` checks in `doctor` and read-only LinkedIn extraction |
  120. | **Help Wanted** | **Regional Job Boards** (GulfTalent, Naukri, StepStone, TokyoDev, LATAM boards) | Open community playbooks — see [`CONTRIBUTING.md`](CONTRIBUTING.md) and [`PLATFORMS.md`](PLATFORMS.md) |
  121. ---
  122. ## Integrations, Compatibility & Secondary Tools
  123. - **AI IDEs & CLI Harnesses (25+ documented):** Antigravity, Claude Code, Claude Desktop, Cursor, Windsurf / Devin Desktop, Codex CLI, Gemini CLI, OpenCode, Kilo Code, Cline, Roo Code, GitHub Copilot, JetBrains Junie / AI Assistant, AWS Kiro, Amazon Q, Trae, Zed, Goose, Amp, Aider, Continue, OpenHands, Warp, Augment, Crush, and Web Chat fallback. Full setup matrix and 6 model tuning guides in [`docs/COMPATIBILITY.md`](docs/COMPATIBILITY.md).
  124. - **Secondary Capability — Local Application Tracking Dashboard:** Inspect your local `logs/applications-log.json` in a zero-server Kanban & channel-yield viewer (`job-hunter dashboard`). Preview the secondary UI with clearly labeled synthetic records at **[Application Tracking Dashboard Demo ↗](https://redasalem.github.io/job-hunter-os/demo-dashboard.html)**.
  125. - **Optional External MCP Plugins:** [`mcp/mcp.template.json`](mcp/mcp.template.json) and [`docs/OPTIONAL_MCP_PLUGINS.md`](docs/OPTIONAL_MCP_PLUGINS.md).
  126. ---
  127. ## Full Skill Catalog & CLI Reference
  128. ### Core Slash Commands
  129. | Command | Workflow | Purpose |
  130. |---|---|---|
  131. | `/demo` | [`.agent/workflows/demo.md`](.agent/workflows/demo.md) | Instant sandbox with 3 synthetic JDs + automatic Job #1 tailored preview |
  132. | `/quick-tailor` | [`.agent/workflows/quick-tailor.md`](.agent/workflows/quick-tailor.md) | Fact-checked HTML/TXT CV tailoring from 1 CV + 1 JD (no browser login) |
  133. | `/setup` | [`.agent/workflows/setup.md`](.agent/workflows/setup.md) | Optional full Career OS onboarding (5 readiness gates + live URL checks) |
  134. | `/daily-run` | [`.agent/workflows/daily-run.md`](.agent/workflows/daily-run.md) | Multi-platform discovery, triage, tailoring & `APPROVE`-gated ATS prep |
  135. | `/apply-url URL` | [`.agent/workflows/apply-url.md`](.agent/workflows/apply-url.md) | Score, tailor, and safely gate a single HTTPS job URL |
  136. | `/tailor-cv URL` | [`.agent/workflows/tailor-cv.md`](.agent/workflows/tailor-cv.md) | Tailored CV + ATS heatmap + `interview-dossier.md` (never submits) |
  137. | `/linkedin-growth` | [`.agent/workflows/linkedin-growth.md`](.agent/workflows/linkedin-growth.md) | Read-only `0–100` Recruiter-SEO audit & copy drafts (`DRAFT_ONLY`, zero browser mutations) |
  138. | `/weekly-report` | [`.agent/workflows/weekly-report.md`](.agent/workflows/weekly-report.md) | Local funnel, rejection pattern, and score calibration insights |
  139. ### CLI Commands (`bin/cli.mjs`)
  140. `audit <cv.md|txt> [--share]` · `init [--global] [--force] [--ide <name>]` · `proof` · `eval [--baseline | --file <json>]` · `pre-submit [--draft <f> | --text <t>]` · `doctor [--strict]` · `status` · `dashboard [--no-open]` · `bundle` · `login` · `unlock` · `backup` · `sync-mcp` · `update`
  141. ### Documentation Index
  142. - [**`AGENTS.md`**](AGENTS.md) & [**`skills/REGISTRY.md`**](skills/REGISTRY.md) — Canonical rules, progressive disclosure index (**48 registered skills**), and 10 guardrails
  143. - [**`docs/SKILL_CATALOG.md`**](docs/SKILL_CATALOG.md) & [**`docs/ARCHITECTURE.md`**](docs/ARCHITECTURE.md) — Complete catalog (`30` core, `4` meta, `7` platform, `7` ATS) and 6-phase pipeline
  144. - [**`docs/execution-safety.md`**](docs/execution-safety.md) & [**`SECURITY.md`**](SECURITY.md) — Deterministic CLI vs. agent workflow controls, privacy boundaries, and signing policy
  145. - [**`HIRED.md`**](HIRED.md) & [**`CHANGELOG.md`**](CHANGELOG.md) — Authentic community outcomes ledger (separated from maintainer origin) & release history
  146. ---
  147. ## Contributing
  148. Contributions are welcome across Markdown playbooks, synthetic ATS fixtures, IDE integrations, and zero-dependency CLI improvements. Three high-impact ways to start (full details, required evidence, and acceptance criteria in [`CONTRIBUTING.md`](CONTRIBUTING.md)):
  149. 1. **Add or improve an ATS synthetic fixture** (`evals/fixtures/forms/` + `evals/RUBRIC.md`).
  150. 2. **Verify or improve a regional job-board playbook** (`skills/platforms/` + `PLATFORMS.md`).
  151. 3. **Validate an AI IDE integration** (`docs/COMPATIBILITY.md` + IDE entrypoint files).
  152. - Report broken ATS/board layouts via [**Playbook Broke**](.github/ISSUE_TEMPLATE/playbook-broke.md).
  153. - Share an authentic job search outcome (`interview` / `contract` / `offer`, anonymous welcome) via [**Share a Job Search Outcome**](https://github.com/redasalem/job-hunter-os/issues/new?template=share-success.yml).
  154. ---
  155. ## License & Attribution
  156. Inspired by open-source patterns from **[`career-ops-hq/career-ops`](https://github.com/career-ops-hq/career-ops)**, **[`obra/superpowers`](https://github.com/obra/superpowers)**, and **[`anthropics/skills`](https://github.com/anthropics/skills)** / **[`agentskills.io`](https://agentskills.io/)**.
  157. Licensed under the [MIT License](LICENSE) — Copyright (c) 2026 Reda Salem (`@redasalem`). See [`NOTICE`](NOTICE) and [`AUTHORS.md`](AUTHORS.md).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.