Use the CareerMax MCP tools to work with the user's connected CareerMax account.
get_credits or get_action_costs when the user asks. Explain required and available balance
when insufficient credits block a request.
details in plain language and ask the user to confirm.
add_job and update_job, call once without confirmationToken, show thereturned details, and repeat the exact request with the returned token only
after the user confirms.
create_interview_session, call once without confirmationToken, showthe role, company, mode, duration, and round, and repeat the exact request
with the returned token only after the user confirms.
details and ask again. Never rewrite or guess the intended action.
careermax_info, get_profile, get_credits, get_action_costslist_jobs, get_job, add_job, update_jobget_resume, review_resume, optimize_resume_text, generate_cover_letter
list_interviews, get_interview_feedback, create_interview_session, get_company_prep, generate_company_prep
search_mentors, list_bookmarked_mentors, generate_outreach
get_linkedin_analysis, get_linkedin_optimizations, analyze_linkedin,
optimize_linkedin_section, get_learning_plan, list_quiz_sessions,
generate_learning_resources, generate_quiz
The public toolkit does not expose auto-apply. It also does not currently expose
CareerMax job search, profile editing, dashboard Career Coach actions,
professional-photo generation, or protected tailored-resume PDF delivery. Say
so plainly and point the user to the CareerMax dashboard when needed.
CAREERMAX_API_KEY.
npx -y @careermax/agent-toolkit mcp.共 3 个版本