CtrlK
BlogDocsLog inGet started
Tessl Logo

careergraph-mentor

Converts a learner’s career background into a structured CareerGraph plan as graph-safe JSON for Neo4j. Use when the user shares their skills, experience, or background and asks what to learn next, how to break into a new role, what skills they are missing, or how to plan a career transition. Also use when asked to map skills to a target career, identify learning paths, suggest projects to build, or recommend resources. Do NOT use for generic career advice, resume writing, or job searching — only when the goal is to produce a structured skill graph plan.

95

1.83x
Quality

93%

Does it follow best practices?

Impact

97%

1.83x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

SKILL.md
Quality
Evals
Security

careergraph-mentor

Converts learner career context into a CareerGraph plan as graph-safe JSON for Neo4j.

Purpose

Produce evidence-based, graph-structured career plans. Do not give shallow advice like “learn React” without evidence, prerequisites, and a concrete proof path. Every recommendation must be expressible as a graph relationship.

Extraction Rules

  • Treat concrete work as stronger evidence than self-reported familiarity.
  • Phrase evidence as observable actions: “Built a static portfolio page”, “Shipped a Flask API”.
  • A skill is only proven when evidence or a project can reasonably demonstrate it.
  • Missing skills must be relevant to the target career and not already proven.
  • Prerequisites must form useful paths from current blockers to target skills or the final goal.
  • Projects should prove multiple missing skills when possible.
  • Resources should teach the next missing skill or a direct prerequisite.
  • Help personas should map to the learner’s current blocker, not the broad career goal.

Output Format

Return JSON only. No prose, no markdown wrapper.

{
  "person": {
    "name": "Demo Learner",
    "targetCareer": "Full-stack Developer"
  },
  "knownEvidence": [
    {
      "description": "Built a static portfolio page",
      "provesSkills": ["HTML", "CSS"]
    }
  ],
  "missingSkills": ["HTTP/API Basics"],
  "careerRequires": ["HTML", "CSS", "HTTP/API Basics"],
  "prerequisites": [["HTTP/API Basics", "Backend Routing"]],
  "projects": [
    {
      "name": "Task Tracker Web App",
      "description": "A CRUD app with login, storage, and deployment.",
      "provesSkills": ["Backend Routing", "Databases", "Authentication"]
    }
  ],
  "resources": [
    {
      "title": "HTTP and APIs for Beginners",
      "type": "interactive guide",
      "url": "https://example.com/http-apis",
      "teachesSkills": ["HTTP/API Basics"]
    }
  ],
  "helpPersonas": [
    {
      "name": "Backend mentor",
      "canHelpWith": ["HTTP/API Basics", "Backend Routing", "Databases"]
    }
  ],
  "nextBestSkill": "HTTP/API Basics",
  "helpPersona": "Backend mentor"
}

Graph Relationship Rules

All recommendations must map to one of these relationships:

  • Evidence -> PROVES -> Skill
  • Skill -> PREREQUISITE_OF -> Skill
  • Project -> PROVES -> Skill
  • Resource -> TEACHES -> Skill
  • HelpPersona -> CAN_HELP_WITH -> Skill

Selecting nextBestSkill

  • Pick from missingSkills only.
  • Prefer prerequisite blockers over advanced framework skills.
  • Prefer projects that prove the most missing skills at once.

Sparse Profiles

If the learner’s profile lacks detail, make conservative assumptions and keep the plan small. Do not fabricate credentials, employers, certificates, or project history.

Repository
alinjfz/CareerMentorGraph
Last updated
Created

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.