# Createrubric > Create a detailed scoring rubric from job posting without performing assessment - Skill: `reggiechan74/createrubric` (Agent Skill) - Install (CLI): `npx skillmds@latest add reggiechan74/createrubric` - Raw SKILL.md: https://api.skillmd.com/api/skills/reggiechan74/createrubric/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: reggiechan74 (https://skillmd.com/u/reggiechan74) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/reggiechan74/createrubric --- ## Configuration Read `.jobops/config.json`. If missing, stop with: > JOBOPS NOT CONFIGURED > Run /jobops:setup to initialize your workspace. Use `config.directories.` for all file paths in this skill. Use `config.preferences.cultural_profile` if this skill generates resume-style content. Use `config.preferences.default_jurisdiction` if this skill has jurisdiction-sensitive logic (crisis/legal skills accept `--jurisdiction=` to override). ## Templates For each template used by this skill, resolve the full path as: {config.templates.base_dir}/{config.templates.active.}/ Templates referenced by this skill: assessment_rubric_framework ## Application Path Resolution This skill writes to a per-application folder. Before writing any output: 1. Parse `{Company}_{Role}_{YYYYMMDD}` from the job-posting filename, or honor `--app=` if supplied. The slug MUST be canonical: a **leading** PascalCase `{Company}` token (matching the `Company_Intelligence/{Company}/` folder so OSINT links), a PascalCase `{Role}` (underscores between words allowed), and a **trailing** compact 8-digit date (`20260519` — no hyphens, no time). Reject leading date/time prefixes such as `2026-04-15_214414_...`; if the source filename carries one, recompose it into canonical form (`{Company}_{Role}_{YYYYMMDD}`) before composing the folder path. 2. Compose the app folder: `{config.directories.applications_root}/{app_slug}/`. 3. Resolve this skill's sub-folder by category: - resume-development (buildresume, provenance-check) → `resume/` - cover-letter (coverletter) → `cover-letter/` - rubric / assessment (createrubric, assessjob, assesscandidate, auditjobposting) → `assessment/` - briefing / interview prep (briefing, interviewprep) → `interview/` 4. If the app folder does not exist, `mkdir -p` it, then copy `{config.directories.job_postings}/{filename}` → `{app_slug}/job_posting.md` so the pinned JD cannot silently change under completed work. Ensure the pinned copy begins with YAML front matter carrying `output_type: job_posting`: if the source JD already has a front-matter block, add the key to it; otherwise wrap a new block (`---` / `output_type: job_posting` / `source_jd: {filename}` / `---`) above the JD body. 5. Exact-slug collisions (same Company+Role+Date) are not auto-suffixed. If the folder already contains the same output type, require the user to pass `--app=`. 6. **Output filename** (fixed; only the sub-folder above is resolved dynamically): write the rubric to `assessment/rubric.md`. ## Your Task Analyze the {{ARG1}} job posting and generate a detailed, reusable scoring rubric that extracts all requirements and creates standardized evaluation criteria. --- ## WORKFLOW ARCHITECTURE ``` Phase 1 (Parallel batch): Load templates + Load job posting (3 parallel reads) Phase 2 (PARALLEL): Domain Research (subagent) while main agent determines role variant Phase 3 (Sequential): Create rubric (synthesizes job posting + domain research) Phase 4 (Sequential): Validate and save rubric ``` **Dependency Rules:** - Phase 2 starts after job posting is loaded (Phase 1) - Domain research subagent and role variant analysis run concurrently - Phase 3 WAITS for domain research results - Phase 4 WAITS for rubric completion --- ## PROGRESS TRACKING (MANDATORY) **Before starting any work**, create all tasks for user visibility: | # | Task Subject | activeForm | |---|-------------|------------| | 1 | Load templates and job posting | Loading templates and job posting | | 2 | Research domain and industry context | Researching domain and industry context | | 3 | Determine role variant and extract requirements | Analyzing role variant and extracting requirements | | 4 | Create 200-point scoring rubric | Creating 200-point scoring rubric | | 5 | Validate rubric completeness | Validating rubric completeness | | 6 | Save rubric | Saving rubric to {applications_root}/{app_slug}/assessment/rubric.md | **Task Update Rules:** - Mark each task `in_progress` BEFORE starting work on it - Mark each task `completed` AFTER finishing it --- ## YAML FRONT MATTER Write the rubric to `{applications_root}/{app_slug}/assessment/rubric.md` and begin the file with: ```yaml --- job_file: {config.directories.job_postings}/{{ARG1}} role: company: role_variant: total_points: 200 generated_by: /createrubric generated_on: output_type: rubric status: final version: 2.0 --- ``` Insert this before the first heading and bump `version` if you update the rubric later. --- ## PHASE 1: LOAD INPUTS (Parallel batch) > **Task:** Mark task 1 `in_progress`. **Read all three files in a single parallel batch using three Read tool calls:** - `{config.templates.base_dir}/{config.templates.active[assessment_rubric_framework]}/assessment_rubric_framework.md` - Master 200-point rubric structure with role variants - `{config.templates.base_dir}/{config.templates.active[evidence_verification_framework]}/evidence_verification_framework.md` - Evidence-based scoring protocols - `{config.directories.job_postings}/{{ARG1}}` (add .md extension if needed) If the job posting doesn't exist in {config.directories.job_postings}/, check the root directory for legacy files. > **Task:** Mark task 1 `completed`. --- ## PHASE 2: PARALLEL ANALYSIS > **Mark tasks 2 and 3 as `in_progress` simultaneously.** > **Dispatch domain research subagent AND begin role variant analysis in the SAME message.** ### 2.1 Research Domain and Industry Context (Task 2 - Subagent) **Dispatch a domain research subagent** to run while you analyze the role variant: ``` Use Task tool with subagent_type=general-purpose, model=sonnet, and prompt: "Research the following for the role of [ROLE TITLE] at [COMPANY NAME]: 1. Industry standards and typical role expectations for this specific position 2. Required vs nice-to-have skills based on current market standards 3. Typical responsibilities and seniority indicators for this role level 4. Company context: culture, values, technology stack, recent developments, size, reputation 5. Current market conditions: salary ranges, demand, competitive landscape 6. Industry-specific terminology, certifications, and best practices 7. What differentiates strong vs average candidates for this type of role Provide a structured research summary organized by these 7 areas. Focus on actionable intelligence that would help calibrate a scoring rubric. Be specific - cite sources and data points where possible." ``` ### 2.2 Determine Role Variant and Extract Requirements (Task 3 - Main agent) **While the domain research subagent runs**, analyze the job posting: #### Determine Role Variant | Variant | Indicators | Weight Adjustment | |---------|------------|-------------------| | **Technical IC** | Hands-on work, no direct reports, technical focus | Skills 30%, Impact 25% | | **People Manager** | Direct reports, team leadership, people development | Skills 20%, Fit 25% | | **Executive** | Strategic scope, P&L authority, cross-functional | Impact 35%, Experience 25% | Document the selected variant in the rubric YAML header. #### Parse Job Posting to Extract Requirements: - **Hard Skills**: All technical competencies with proficiency levels required - **Experience Requirements**: Relevance, recency, and domain specifics (NOT years-based scoring) - **Impact Evidence**: Demonstrated achievements, outcomes, and metrics - **Credentials**: Required and preferred education/certifications - **Fit & Readiness**: Communication, values alignment, and role readiness indicators > **Task:** Mark task 3 `completed` when extraction is done. > **Task:** Mark task 2 `completed` when domain research subagent returns. --- ## PHASE 3: CREATE RUBRIC (Sequential - needs domain research) > **Task:** Mark task 4 `in_progress`. > **Prerequisite:** Domain research (task 2) must be `completed`. ### 3.1 Synthesize Job Posting + Domain Research Use domain research findings to calibrate rubric thresholds: - Set proficiency levels appropriate to industry standards - Weight skills based on market reality, not just posting emphasis - Identify implicit requirements the posting may not state - Calibrate experience expectations to seniority level ### 3.2 Create Job-Specific Rubric Following Template Structure **MANDATORY DETAILED SCORING REQUIREMENT**: You MUST create a comprehensive **200-point scoring rubric** following the EXACT structure defined in the assessment rubric framework template. This is NON-NEGOTIABLE. **CRITICAL**: Use the assessment rubric framework template as your MANDATORY reference template. The rubric structure is fixed - you must maintain: 1. **Five Main Categories** (default point allocations - adjust based on role variant): - Skills Inventory (50 points default / 25%) - Experience Relevance (40 points default / 20%) - Demonstrated Impact (60 points default / 30%) - Credentials (20 points default / 10%) - Fit & Readiness (30 points default / 15%) 2. **Role Variant Weight Adjustments**: | Variant | Skills | Experience | Impact | Credentials | Fit | |---------|--------|------------|--------|-------------|-----| | Technical IC | 30% (60) | 20% (40) | 25% (50) | 10% (20) | 15% (30) | | People Manager | 20% (40) | 20% (40) | 20% (40) | 15% (30) | 25% (50) | | Executive | 15% (30) | 25% (50) | 35% (70) | 10% (20) | 15% (30) | 3. **Proficiency-Based Skill Scoring** (7-level scale, NOT years-based): - 6 - Expert: Can teach/design novel solutions; recognized authority - 5 - Advanced: Solves complex problems independently; mentors others - 4 - Proficient: Handles standard work with minimal guidance - 3 - Competent: Performs with occasional supervision - 2 - Developing: Requires regular guidance; growing capability - 1 - Novice: Basic understanding; needs significant support - 0 - None: No demonstrated capability 4. **Mandatory Rubric Components**: - **Alignment Statement**: Construct definition explaining what the rubric measures - **Critical Barriers Table**: Minimum thresholds with consequences for failure - **Anchor Examples**: Concrete examples for each scoring level - **Confidence Flagging**: Guidance for marking low-confidence scores - **Weight Justification Table**: Rationale for category weight allocations 5. **Customize Job-Specific Content**: - Replace `[bracketed placeholders]` with actual job requirements - Extract specific skills, technologies, and responsibilities from job posting - Define role-specific thresholds and criteria (scope, budget size, team size, etc.) - Add company-specific values and cultural indicators - Include industry-specific success metrics and KPIs 6. **Maintain Template Components**: - Overall assessment ranges for each section - Evaluation frameworks with specific metrics - Evidence-based scoring verification protocols - Usage guidelines and quality control checklists - Scoring interpretation guidelines > **Task:** Mark task 4 `completed`. --- ## PHASE 4: VALIDATE AND SAVE ### 4.1 Validate Rubric Completeness > **Task:** Mark task 5 `in_progress`. **ENFORCEMENT CHECK**: After creating the rubric, verify: - Role variant selected and documented in YAML header - Point allocations match selected variant (total = 200) - Five categories present: Skills, Experience, Impact, Credentials, Fit - Alignment Statement with construct definition included - Critical Barriers table with thresholds and consequences - Skills use 7-level proficiency scale (0-6), NOT years-based - No redundancy (years only in Experience, achievements only in Impact) - Anchor examples for each scoring level - Confidence flagging guidance included - Weight justification table present - Evidence verification framework is included **IF ANY SECTION LACKS REQUIRED COMPONENTS, THE RUBRIC IS INCOMPLETE AND MUST BE REGENERATED** > **Task:** Mark task 5 `completed`. ### 4.2 Save the Rubric > **Task:** Mark task 6 `in_progress`. Save the generated rubric to: `{applications_root}/{app_slug}/assessment/rubric.md` Provide a summary of: - Template compliance confirmation - Total skills/requirements identified - Key differentiators for this role - Notable industry-specific requirements - Job-specific customizations made to template - Recommended usage guidelines > **Task:** Mark task 6 `completed`. --- ## Important Notes - **Template-Based Approach**: Every rubric MUST follow the assessment rubric framework template structure exactly - **200-Point System**: Total score is 200 points (normalized to percentage for comparison) - **5 Categories**: Skills Inventory, Experience Relevance, Demonstrated Impact, Credentials, Fit & Readiness - **Role Variant Selection**: Mandatory classification as Technical IC, People Manager, or Executive - **Proficiency-Based Scoring**: Skills use 7-level proficiency scale (0-6), NOT years of experience - **No Redundancy**: Years of experience only in Experience category; achievements only in Impact category - **Domain Research**: Domain research informs rubric calibration - thresholds and weights should reflect industry reality - **Complete Scoring Levels**: All sections must include anchor examples for each scoring level - **Critical Barriers**: Must define minimum thresholds with consequences for failing to meet them - **Evidence Framework**: Include evidence verification protocols from the evidence verification framework template - **Reusable Rubrics**: Save rubric for consistent evaluation across multiple candidates for the same position - **No Shortcuts**: Generated rubric must be as detailed as the template - simplified versions violate requirements ## CRITICAL ENFORCEMENT RULES **ABSOLUTE REQUIREMENTS FOR EVERY RUBRIC:** Must follow the assessment rubric framework template structure with: 1. **200-POINT TOTAL**: All rubrics use 200-point scale (normalized to percentage for comparison) 2. **5 CATEGORIES ONLY**: Skills Inventory, Experience Relevance, Demonstrated Impact, Credentials, Fit & Readiness 3. **ROLE VARIANT**: Must select and document Technical IC, People Manager, or Executive variant 4. **PROFICIENCY SCALE**: Skills use 7-level proficiency scale (0-6), NOT years-based scoring 5. **NO REDUNDANCY**: Years only in Experience category; achievements only in Impact category 6. **ALIGNMENT STATEMENT**: Must include construct definition explaining what the rubric measures 7. **CRITICAL BARRIERS**: Must define minimum thresholds with consequences for failing 8. **ANCHOR EXAMPLES**: Must provide concrete examples for each scoring level 9. **CONFIDENCE FLAGGING**: Must include guidance for marking low-confidence scores 10. **WEIGHT JUSTIFICATION**: Must include table explaining category weight allocations 11. **EVIDENCE PROTOCOLS**: Must include verification framework from the evidence verification framework template **VIOLATION CONSEQUENCES:** - Any rubric not following template structure is INCOMPLETE and violates the command specification - You must regenerate the rubric if any section deviates from template - Simplified or abbreviated rubrics are NOT ACCEPTABLE - The rubric must match the template structure exactly - no exceptions **VERIFICATION CHECKLIST - BEFORE SAVING ANY RUBRIC:** - [ ] Role variant selected and documented in YAML header - [ ] Point allocations match selected variant (total = 200) - [ ] Five categories present with correct weight distribution - [ ] Alignment Statement with construct definition included - [ ] Critical Barriers table with thresholds and consequences - [ ] Skills use 7-level proficiency scale (0-6), NOT years-based - [ ] No redundancy between categories - [ ] Anchor examples for each scoring level - [ ] Confidence flagging guidance included - [ ] Weight justification table present - [ ] Evidence verification protocols included - [ ] Job-specific content customized within template structure