CareerProof MCP

Add a plain text artifact to teach a custom eval model. Use for free-form descriptions of what good/bad looks like, rubric notes, or evaluation criteria text. artifact_type: cv_with_notes, template, free_text, jd. label: strong, weak, or mixed. model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

Send a message to Atlas Advisor for lightweight hiring advice (2 credits). Faster and cheaper than atlas_chat, no tool use -- best for general hiring questions. Returns AI response text and a conversation_id. Omit conversation_id to start a new conversation; include it to continue a thread.

Send a message to Atlas AI assistant (3 credits). Provides workforce intelligence, candidate insights, and hiring strategy advice with live tool use (salary benchmarks, talent supply/demand, competitor intel). Returns AI response text and a conversation_id. Omit conversation_id to start a new conversation; include it to continue an existing thread. Optionally pass context_id from atlas_list_contexts for context-aware responses. Note: may take 2-3 minutes for complex queries with multiple tool calls.

Clear all rubric overrides for a custom eval model, reverting to the AI-inferred rubric. model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

Create a new hiring context (company profile) for organizing candidates and JDs. Returns a context object with 'id'. Use this id as context_id in atlas_upload_candidate, atlas_create_jd, atlas_list_candidates, and all analysis/batch/utility tools. This is the first step in the Atlas workflow. Free.

Create a new custom evaluation model for a hiring context. Returns the model object with 'id'. Use this id as model_id in atlas_upload_custom_eval_artifact, atlas_add_custom_eval_text_artifact, atlas_infer_custom_eval_rubric, atlas_start_custom_eval_batch, and atlas_start_custom_eval_inference. Requires client_context_id from atlas_create_context or atlas_list_contexts. Free.

Create a job description from text within a hiring context. Returns a JD object with 'id' and stored content. Use JD content as jd_text in atlas_fit_match, atlas_fit_rank, atlas_start_jd_fit_batch, and atlas_start_jd_analysis. Requires context_id from atlas_create_context or atlas_list_contexts. Free.

Delete a custom evaluation model. This removes the model and all associated artifacts and rubrics. model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

Download a completed report as PDF. Returns base64-encoded PDF content. Confirm report status='completed' via atlas_get_report(report_id) first. report_id from atlas_start_report response or atlas_list_reports. Free.

Enhanced FIT match with knowledge base augmentation for deeper analysis (8 credits). Same inputs as atlas_start_fit_match but augments with industry KB data for richer insights. Returns fit score (0-100), strengths, gaps, recommendations, and KB-sourced market context. Synchronous. Requires context_id, candidate_id, and jd_text -- same sources as atlas_start_fit_match.

Generate interview questions tailored to a candidate's profile (8 credits). Returns an array of interview questions with rationale. Supports pressure levels: supportive, standard, aggressive. Optionally pass jd_text for role-targeted questions. After the interview, use atlas_interview_followup for follow-up probing. Synchronous. Requires context_id from atlas_list_contexts and candidate_id from atlas_list_candidates.

Retrieve a completed analysis result by analysis ID. Returns scores, competency breakdown, and recommendations. analysis_id comes from atlas_start_gem_analysis response or atlas_list_analyses. Only works after analysis is completed -- check with careerproof_task_status first. Free.

Get comprehensive talent analytics dashboard data for a hiring context (candidate stats, analysis coverage, fit score distributions). Requires context_id from atlas_create_context or atlas_list_contexts. Read-only summary, no downstream chaining needed. Free.

Get all results from a completed batch GEM job. Returns an array of per-candidate analysis results with scores and competency breakdowns. Check atlas_get_batch_gem_status(job_id) first to confirm status='completed'. job_id from atlas_start_batch_gem response. Free.

Check progress of a batch GEM job. Returns status (pending/running/completed/failed) and progress percentage. When status='completed', fetch results with atlas_get_batch_gem_results(job_id). job_id comes from atlas_start_batch_gem response. Free.

Get full candidate detail including parsed CV content and parse status. Use this to verify CV parsing is complete (status='completed') before starting analysis. Requires context_id and candidate_id from atlas_upload_candidate or atlas_list_candidates. Free.

Get all results from a completed custom eval batch. Returns per-candidate evaluation scores and dimension breakdowns. Confirm status='completed' via atlas_get_custom_eval_batch_status(batch_id) first. batch_id from atlas_start_custom_eval_batch response. Free.

Check progress of a custom evaluation batch. Returns status and progress. When status='completed', fetch results with atlas_get_custom_eval_batch_results(batch_id). batch_id from atlas_start_custom_eval_batch response. Free.

Get full detail for a custom evaluation model including dimensions, weights, artifacts summary, and rubric status. model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

Get the current rubric for a custom eval model. Returns evaluation dimensions, scoring criteria, and any user overrides. model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

Get assessment results for a dialogue session. Returns interview quality scores, communication assessment, and recommendations. Confirm assessment is completed via careerproof_task_status first. session_id from the dialogue session used in atlas_start_dialogue_assessment. Free.

Check progress of a JD-FIT batch job. Returns status and progress percentage. When complete, fetch results with atlas_get_jd_fit_results(context_id). batch_id comes from atlas_start_jd_fit_batch response. Free.

Get JD-FIT results for a context. Returns ranked candidates with fit scores, strengths, and gaps. Confirm batch is complete via atlas_get_jd_fit_batch_status first. context_id from atlas_create_context or atlas_list_contexts. Free.

Get report status and metadata. Returns status (pending/generating/completed/failed), title, type, and summary. When status='completed', download the PDF with atlas_download_report(report_id). report_id from atlas_start_report response or atlas_list_reports. Free.

Infer an evaluation rubric from the model's artifacts. AI analyzes all uploaded artifacts and generates evaluation dimensions with scoring criteria. This is a synchronous AI call that may take 30-60 seconds. model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

Generate a follow-up question based on a candidate's answer to an interview question (1 credit). Returns a follow-up question with probing rationale. Use after atlas_generate_interview -- pass the original_question from that tool's output and the candidate's response. Synchronous. Requires context_id and candidate_id.

List all analyses for a candidate with optional type filtering. Returns an array with each analysis's id, type, status, and scores. Use the analysis id with atlas_get_analysis to retrieve full details. Requires context_id and candidate_id from atlas_list_candidates. Free.

List all candidates in a hiring context. Returns an array of candidate objects, each with an 'id' field. Use candidate id as candidate_id in atlas_start_gem_analysis, atlas_fit_match, atlas_fit_rank, atlas_generate_interview, and batch tools. Requires context_id from atlas_create_context or atlas_list_contexts. Free.

List all hiring contexts for the authenticated user. Returns an array of context objects, each with an 'id' field. Use the id as context_id in atlas_upload_candidate, atlas_create_jd, atlas_list_candidates, analysis, batch, and utility tools. Free.

List all artifacts uploaded to a custom eval model. Returns artifact objects with id, type, label, notes, and upload status. model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

List all custom evaluation models for the authenticated user. Returns an array of model objects with id, name, description, and status. Use model id in artifact, rubric, and evaluation tools. Free.

List all job descriptions for a hiring context. Returns an array of JD objects with id, title, and content. Use JD content as jd_text in atlas_fit_match, atlas_fit_rank, and atlas_start_jd_fit_batch. Requires context_id from atlas_create_context or atlas_list_contexts. Free.

List all generated reports with status and summary info. Returns an array of report objects with id, report_type, status, title, and summary. Use the report id with atlas_get_report for details or atlas_download_report to download completed PDFs. Free.

List all 16 available report types with descriptions, categories, credit costs, and required/optional input fields. Call this BEFORE atlas_start_report to discover valid report_type values and the exact inputs each type requires. Categories: RECRUITMENT (7 types), WORKFORCE_PLANNING (6 types), STRATEGIC_HR (3 types). Free.

Set rubric overrides for a custom eval model. Overrides let you customize the AI-inferred rubric (adjust weights, rename dimensions, add scoring criteria) without re-inferring. Replaces any existing overrides. model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

Start a batch GEM analysis across multiple candidates (10 credits per candidate). Returns a job_id. Poll with atlas_get_batch_gem_status(job_id) until status='completed', then fetch results with atlas_get_batch_gem_results(job_id). Requires context_id from atlas_list_contexts and candidate_ids from atlas_list_candidates. All candidate CVs must be fully parsed.

Start batch evaluation of multiple candidates using a custom evaluation model (5 credits per candidate). Returns a batch_id. Poll with atlas_get_custom_eval_batch_status(batch_id) until status='completed', then fetch with atlas_get_custom_eval_batch_results(batch_id). Requires context_id from atlas_list_contexts, candidate_ids from atlas_list_candidates, and custom_model_id from the Atlas dashboard.

Start async inference on a custom evaluation model to auto-generate evaluation dimensions (5 credits). Returns a task_id. Poll with careerproof_task_status(task_id) until status='completed', then fetch results with careerproof_task_result(task_id, result_type='custom_eval_inference', resource_id=model_id). model_id must be from an existing custom eval model.

Start async assessment of dialogue/interview transcripts in a session (10 credits). Returns a task_id. Poll with careerproof_task_status(task_id) until status='completed', then fetch results with atlas_get_dialogue_results(session_id). session_id must be from an existing dialogue session.

Start an async fit match for a single candidate against a job description (8 credits). Returns task_id and analysis_id. Poll with careerproof_task_status, then fetch result with careerproof_task_result (result_type='fit_match'). Requires context_id from atlas_list_contexts, candidate_id from atlas_list_candidates, and jd_text (from atlas_list_jds or raw text). For multiple candidates, use atlas_start_fit_rank or atlas_start_jd_fit_batch instead.

Start an async rank of multiple candidates against a job description (8 credits). Returns task_id and analysis_id. Poll with careerproof_task_status, then fetch result with careerproof_task_result (result_type='fit_rank'). Requires context_id from atlas_list_contexts, candidate_ids from atlas_list_candidates (minimum 2), and jd_text. For async batch processing with more detail, use atlas_start_jd_fit_batch instead.

Start an async GEM (10-factor competency) analysis on a candidate. Returns a task_id and analysis_id. Poll with careerproof_task_status(task_id) until status='completed', then fetch results with atlas_get_analysis(analysis_id) or careerproof_task_result(task_id, result_type='analysis', resource_id=analysis_id). Candidate CV must be fully parsed first -- verify with atlas_get_candidate. Types: gem_full (10 cr), gem_lite (5 cr), career_path (5 cr).

Analyze a job description text for market competitiveness, clarity issues, bias detection, salary benchmarking, and skills extraction (5 credits). Synchronous -- returns results directly (no polling needed). Does not require a context or candidate; accepts raw JD text. Use results to refine JDs before running atlas_start_jd_fit_batch.

Start a batch JD-FIT analysis: match multiple candidates against a job description (3 credits per candidate). Returns a batch_id. Poll with atlas_get_jd_fit_batch_status(context_id, batch_id) until complete, then fetch with atlas_get_jd_fit_results(context_id). If jd_content is omitted, uses the context's active JD. Requires context_id from atlas_list_contexts and candidate_ids from atlas_list_candidates.

Start async generation of a research report (15 credits). IMPORTANT: Call atlas_list_report_types first to get valid report_type values and required inputs. Returns report_id and task_id. Poll with careerproof_task_status(task_id) or atlas_get_report(report_id) until status='completed', then download PDF with atlas_download_report(report_id).

Update a hiring context's details (name, industry, size, hiring context). Requires context_id from atlas_create_context or atlas_list_contexts. Returns the updated context object. Free.

Update a custom evaluation model's metadata (name, description, inferred dimensions, weights). model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Returns the updated model object. Free.

Upload a candidate CV (PDF/DOCX) to a hiring context. CV parsing starts asynchronously. Returns a candidate object with 'id' -- use this as candidate_id in atlas_start_gem_analysis, atlas_fit_match, atlas_generate_interview, and batch tools. Use atlas_get_candidate to check if parsing is complete before running analysis. Requires context_id from atlas_create_context or atlas_list_contexts. Free.

Upload a file artifact (CV, JD, template) to teach a custom eval model. Accepts PDF/DOCX as base64. artifact_type: cv_with_notes, template, free_text, jd. label: strong, weak, or mixed (how this artifact exemplifies quality). model_id from atlas_create_custom_eval_model or atlas_list_custom_eval_models. Free.

Retrieve the full result of a completed async task. Call careerproof_task_status first to confirm status='completed'. The resource_id varies by result_type: 'analysis' needs analysis_id (from atlas_start_gem_analysis), 'batch_gem' needs job_id (from atlas_start_batch_gem), 'report' needs report_id (from atlas_start_report), 'custom_eval_inference' needs model_id, 'custom_eval_batch' needs batch_id, 'dialogue_assessment' needs session_id. 'jd_fit_batch' requires context_id instead of resource_id. Free.

Check the status of any async task. Returns status (pending/running/completed/failed), progress percentage, and current stage. When status='completed', fetch full results with careerproof_task_result. Recommended polling interval: every 5-10 seconds. The task_id comes from any async start tool (atlas_start_gem_analysis, atlas_start_batch_gem, atlas_start_report, atlas_start_dialogue_assessment, atlas_start_custom_eval_inference, etc.). Free.

Run market positioning analysis on a CV version (5 credits, takes 20-30s). Returns positioning snapshot, detected narrative lens, recruiter inference, mixed signal flags, and a session_id. This is step 1 of the 3-step positioning pipeline: analyze_positioning -> ceevee_get_opportunities(lens) -> ceevee_confirm_lens. Pass the returned session_id to subsequent steps. cv_version_id from ceevee_upload_cv or ceevee_list_versions.

Send a message to CeeVee AI assistant for CV optimization guidance (2 credits). Requires a cv_version_id (use ceevee_upload_cv or ceevee_list_versions to get one). Returns AI response with optional edit suggestions, source citations, and a conversation_id. Omit conversation_id to start a new conversation; include it to continue a thread.

Confirm a narrative lens and generate targeted CV edits with trade-offs (5 credits, takes 20-30s). Returns an array of section edits with before/after text, trade-off notes, and optionally clean + review PDF download URLs. This is step 3 (final step) of the positioning pipeline. Pass confirmed_lens from ceevee_analyze_positioning, and optionally positioning_snapshot, detected_lens_full, recruiter_inference, selected_opportunities from prior steps for richer edits. Use ceevee_explain_change to understand any specific edit.

Download a completed CeeVee report as PDF. Returns base64-encoded PDF content. Only call after ceevee_get_report confirms status='completed'. report_id from ceevee_generate_report response or ceevee_list_reports. Free.

Get a detailed explanation of a specific edit from a positioning review (1 credit). The change_id comes from the edits array returned by ceevee_confirm_lens or ceevee_full_review. Returns a detailed rationale for the recommended change. cv_version_id from ceevee_upload_cv or ceevee_list_versions.

Run autonomous full CV review in a single call: positioning + lens detection + targeted edits + optional opportunities (10 credits, takes 30-60s). Alternative to the 3-step pipeline (ceevee_analyze_positioning -> ceevee_get_opportunities -> ceevee_confirm_lens). Returns positioning snapshot, detected lens, targeted edits with trade-offs, and optional opportunities. cv_version_id from ceevee_upload_cv or ceevee_list_versions.

Generate a CeeVee career-intel report asynchronously (15 credits, takes 2-3 min). Returns report_id and status. POLLING: Call ceevee_get_report(report_id) every 30 seconds, up to 40 times (20 min max). If status='completed', download PDF with ceevee_download_report(report_id). If status='failed', relay the error_message to the user. If still processing after 40 polls, stop and inform the user with the report_id so they can check later. Call ceevee_list_report_types first to discover valid report_type values and required inputs. Report categories: Compensation Benchmark, Role Evolution, Offer Comparison, AI Displacement Risk, Pivot Feasibility, Credential ROI, Skill Decay Risk, Rate Card, Career Gap Narrative, Interview Prep, Employer Red Flag, Industry Switch, Relocation Impact, Startup vs Corporate, Learning Path, Board Readiness, Fractional Leadership.

Get career pivot opportunities based on the CV and a selected narrative lens (3 credits). Returns 2-4 opportunities with rationale, CV signals, and market context. This is step 2 of the positioning pipeline (after ceevee_analyze_positioning). The 'lens' value should come from ceevee_analyze_positioning output (e.g. 'Technical Leader', 'Scale-up Builder'). Pass the same session_id from step 1. Next step: ceevee_confirm_lens with selected opportunities.

Get full positioning session detail including analysis, lens, opportunities, targeted edits, and PDF download URLs. session_id from ceevee_list_positioning_sessions or from ceevee_analyze_positioning response. Contains everything generated across the analyze -> opportunities -> confirm-lens pipeline. Free.

Get report status and metadata (without PDF). Returns status (pending/processing/completed/failed), title, type, inputs, and summary. This is the polling tool for ceevee_generate_report — call every 30 seconds, up to 40 times (20 min max). When status='completed', download PDF with ceevee_download_report(report_id). If status='failed', relay error_message. If still processing after 40 polls, stop and give the user the report_id to check later. Free.

Get detailed CV version including structured content, sections, word count, and audience profile. cv_version_id from ceevee_upload_cv or ceevee_list_versions. Use to inspect CV content before running analysis tools. Free.

List all positioning sessions (market analysis through lens selection to targeted edits). Returns an array of session objects with id, status, cv_version_id, and created_at. Use the session id with ceevee_get_positioning_session for full details including analysis results, edits, and PDFs. Free.

List user's generated CeeVee reports with pagination and optional status filter. Returns report summaries with id, report_type, status, title, and summary. Use the report id with ceevee_get_report for details or ceevee_download_report to download completed PDFs. Free.

List all available CeeVee career-intel report types with descriptions, required/optional input fields, and credit costs. Call this BEFORE ceevee_generate_report to discover valid report_type values and the exact inputs each type requires. Categories include Compensation Benchmark, Role Evolution, Offer Comparison, AI Displacement Risk, Pivot Feasibility, Credential ROI, Skill Decay Risk, Rate Card, Career Gap Narrative, Interview Prep, Employer Red Flag, Industry Switch, Relocation Impact, Startup vs Corporate, Learning Path, Board Readiness, and Fractional Leadership. Free.

List all CV versions for the authenticated user. Returns an array of version objects with id, filename, created_at, and main_version flag. Use the version id as cv_version_id in ceevee_analyze_positioning, ceevee_full_review, ceevee_confirm_lens, and ceevee_chat. Free.

Upload a CV file (PDF/DOCX). Parses the document, creates a CV version, and returns structured sections with word count. Returns a version object with 'id' -- use this as cv_version_id in ceevee_analyze_positioning, ceevee_full_review, ceevee_confirm_lens, ceevee_chat, and ceevee_explain_change. This is the first step in the CeeVee workflow. Free.