Menu
Accelerate debugging by providing diagnostic context that enables agent reasoning through problem spaces. Use when encountering errors, unexpected behavior, or when standard debugging produces circular results.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Create, update, close, and report on Jira tickets. Reads JIRA_URL and JIRA_TOKEN from environment variables or .env file. Provides comprehensive guidance for Jira ticket lifecycle management, API integration, and automation workflows.
AI-powered clarity engine for job seekers and founders. Not a resume fixer — a thinking partner. Helps you find your north star sentence: who you are, what you do, and why it matters to someone else. Produces one core version and targeted variations for different audiences. Use before interviews, investor conversations, networking events, or any moment where you need to introduce yourself under pressure.
Review code for bugs, security issues, and improvements. Use when the user asks to: (1) review code or a diff, (2) check code quality, (3) find bugs or issues, (4) audit for security problems. Trigger on phrases like 'review this', 'check my code', 'any issues with this', 'code review'.
Full-lifecycle career advisor — self-discovery through career growth. Use when the user wants to: (1) discover identity, values, ikigai, or skills inventory, (2) define dream job or analyze career gaps, (3) close gaps via learning, certifications, portfolio, or personal branding, (4) build narrative hats or role-specific positioning, (5) search for jobs, generate resumes, or prepare for interviews, (6) onboard into a new role, track achievements, or generate self-reviews, (7) backfill past roles, run weekly recall, or check career readiness. Proactively nudges users to document achievements NOW — not when they need a resume. Consolidates the former career-resume, career-tracker, portfolio-strategy, and github-profile skills.
Synthesizes professional accomplishments into high-converting landing page content. Triggers when a user wants to build a landing page, launch a consulting service, productize their career expertise, create a service page, or monetize their skills. Also triggers on: "freelance", "consulting offer", "offer page", "service offering", "turn my experience into a business". Uses career-advisor data as the foundation.
Query Google NotebookLM notebooks directly from Claude Code for source-grounded, citation-backed answers. Use when user mentions NotebookLM, shares a notebooklm.google.com URL, asks to query their notebooks/docs, or says "ask my NotebookLM", "check my docs", "query my notebook". Provides browser automation, notebook library management, and persistent auth. Drastically reduced hallucinations through document-only responses from Gemini.
| name | context-aware-debugging |
| description | Accelerate debugging by providing diagnostic context that enables agent reasoning through problem spaces. Use when encountering errors, unexpected behavior, or when standard debugging produces circular results. |
Errors are symptoms. Context reveals causes. Feed the agent what it needs to reason, not just react.
Every debug prompt needs three elements:
"Error: Cannot read property 'map' of undefined in UserList.jsx line 42.
Expected: render array of users from API.
The API returns { data: { users: [...] } }"
"Authentication flow breaks after login redirect.
- Error: 401 on /api/profile after successful /api/login
- Expected: JWT in cookie should authorize subsequent requests
- Recent change: Added sameSite cookie attribute
- Relevant files: auth.middleware.js, login.controller.js
- Browser: Cookie IS being set (verified in DevTools)"
Before prompting the agent:
| Weak | Strong |
|---|---|
| "It doesn't work" | "Returns null instead of user object when..." |
| "Fix this error" + stack trace | Stack trace + expected behavior + recent changes |
| "Why is this broken?" | "This breaks when X but works when Y" |
When stuck in a loop: "Explain three different hypotheses for why this could fail, then tell me what evidence would confirm or eliminate each."
The agent becomes a reasoning partner, not just a fix-generator.