job-hunt-copilot MCP — Cas d'usage, Installer & Démo en direct

Pourquoi l'utiliser

Fonctionnalités clés

Project library as a single source of truth
Per-role resume tailoring without rewriting history
Pitch script generation aligned to role signals
Mock interview mode using role-specific questions
Keeps honesty — edits framing, not facts

Démo en direct

Aperçu en pratique

job-hunt-copilot-skill.replay ▶ prêt

0/0

Installer

Choisissez votre client

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "job-hunt-copilot-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/spontaneousai/job-hunt-copilot",
        "~/.claude/skills/job-hunt-copilot"
      ],
      "_inferred": true
    }
  }
}

Ouvrez Claude Desktop → Settings → Developer → Edit Config. Redémarrez après avoir enregistré.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "job-hunt-copilot-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/spontaneousai/job-hunt-copilot",
        "~/.claude/skills/job-hunt-copilot"
      ],
      "_inferred": true
    }
  }
}

Cursor utilise le même schéma mcpServers que Claude Desktop. La config projet l'emporte sur la globale.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "job-hunt-copilot-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/spontaneousai/job-hunt-copilot",
        "~/.claude/skills/job-hunt-copilot"
      ],
      "_inferred": true
    }
  }
}

Cliquez sur l'icône MCP Servers dans la barre latérale Cline, puis "Edit Configuration".

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "job-hunt-copilot-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/spontaneousai/job-hunt-copilot",
        "~/.claude/skills/job-hunt-copilot"
      ],
      "_inferred": true
    }
  }
}

Même format que Claude Desktop. Redémarrez Windsurf pour appliquer.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "job-hunt-copilot-skill",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/spontaneousai/job-hunt-copilot",
        "~/.claude/skills/job-hunt-copilot"
      ]
    }
  ]
}

Continue utilise un tableau d'objets serveur plutôt qu'une map.

~/.config/zed/settings.json

{
  "context_servers": {
    "job-hunt-copilot-skill": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/spontaneousai/job-hunt-copilot",
          "~/.claude/skills/job-hunt-copilot"
        ]
      }
    }
  }
}

Ajoutez dans context_servers. Zed recharge à chaud à la sauvegarde.

claude mcp add job-hunt-copilot-skill -- git clone https://github.com/spontaneousai/job-hunt-copilot ~/.claude/skills/job-hunt-copilot

Une seule ligne. Vérifiez avec claude mcp list. Supprimez avec claude mcp remove.

Cas d'usage

Usages concrets : job-hunt-copilot

Tailor your resume to a specific JD without starting over

👤 Job seekers who send 5+ applications a week ⏱ ~15 min beginner

Quand l'utiliser : You found a role, you have a base resume, you don't want to rewrite from scratch.

Prérequis

Project library set up — Follow SKILL.md to populate projects.md with detailed project notes once
Skill installed — git clone https://github.com/spontaneousai/job-hunt-copilot ~/.claude/skills/job-hunt-copilot

Déroulement

Paste the JD

Use job-hunt-copilot. Here's the JD: [...]. Tailor my resume to it using my project library.✓ Copié

→ Resume with bullets reshuffled and reworded to match JD keywords — without inventing experience
Verify honesty

Show me any bullet where the framing is a stretch from my actual project notes.✓ Copié

→ Flags for review, not a clean 'all good'
Export

Save as resume-acme-senior-pm.md and also convert to PDF.✓ Copié

→ File saved with clean filename

Résultat : A tailored resume per role in ~10 minutes, same facts, different emphasis.

Pièges

Hallucinated accomplishments — Always use the honesty-check step; your project library is the source of truth

Combiner avec : filesystem

Generate a 2-minute 'tell me about yourself' pitch for a role

👤 Interviewees preparing for first-round screens ⏱ ~20 min beginner

Quand l'utiliser : You have a screen in 24 hours.

Déroulement

Align pitch to JD

Use job-hunt-copilot. Write my 2-min pitch for this role [JD]. Emphasize the projects that map to their top 3 requirements.✓ Copié

→ Script with natural flow, not a resume readout
Practice variants

Now give me a 45-second version and a 5-minute version.✓ Copié

→ Three lengths, same core

Résultat : A pitch that sounds like you, tuned to the role.

Pièges

Sounding robotic — Read aloud; trim anything that feels LinkedIn-flavored

Run a mock interview for a specific role

👤 Candidates with a real interview soon ⏱ ~45 min intermediate

Quand l'utiliser : You want to rehearse the likely questions.

Déroulement

Set up the role context

Use job-hunt-copilot. Act as a hiring manager for this role [JD] doing a behavioral interview. Ask me questions one at a time.✓ Copié

→ One question, waits for answer, then probes
Get feedback

After we finish, critique my answers — structure, specificity, signal.✓ Copié

→ Per-answer feedback with improvement hooks

Résultat : A rehearsed set of stories for the interview.

Pièges

Mock becomes chummy and doesn't probe — Tell Claude to be skeptical and ask follow-up 'why' questions

Combinaisons

Associez-le à d'autres MCPs pour un effet X10

job-hunt-copilot-skill + filesystem

Keep a structured folder per role: JD, resume version, pitch, notes

For every new role, create jobs/<company>/<role>/ with JD, resume, pitch, and interview notes.✓ Copié

Outils

Ce que ce MCP expose

Outil	Entrées	Quand appeler
tailor_resume	JD, base resume, project library	Per role application
pitch_script	JD, focus projects	Before a screen
mock_interview	role type, rounds	Rehearsal
honesty_check	resume draft	After tailoring, before sending

Coût et limites

Coût d'exécution

Quota d'API: None
Tokens par appel: Modest per role; the library doc is the biggest token cost
Monétaire: Free
Astuce: Keep projects.md well-organized — smaller, dense entries beat long narratives.

Sécurité

Permissions, secrets, portée

Stockage des identifiants : None

Sortie de données : Project library and JDs sent to Claude API

Don't put current-employer confidential information into your project library. Your own accomplishments are fair game; their trade secrets aren't.

Dépannage

Erreurs courantes et correctifs

Tailored resume invents claims

Your project library may be too sparse; expand it so tailoring has real material to pull from

Vérifier : Open projects.md and check each entry has metrics and details

Pitch sounds LinkedIn-y

Prompt for 'natural spoken words, not buzzwords'; read aloud

Alternatives

job-hunt-copilot vs autres

Alternative	Quand l'utiliser	Compromis
ChatGPT + manual tailoring	One-off applications	No persistent library; you repeat yourself
Teal / JobScan	Keyword matching only	No narrative tailoring

Plus

Ressources

📖 Lire le README officiel sur GitHub

🐙 Voir les issues ouvertes

🔍 Parcourir les 400+ serveurs MCP et Skills