Career
Career management including resume coaching, job evaluation, and career pipeline tools.
When to Invoke
User can say things like:
- "Run a resume review on [file or role]"
- "Review my resume for [formatting / design / content / this job]"
- "Score this job" / "Evaluate this job posting" / "Should I apply to this role?"
- "Help with rubric development" or "Evaluate this resume adaptation"
- "Develop the content set from remaining convos" or "Pull more from conversations"
- "Audit the resume chain" or "Can I trust this pipeline?"
Route to the right workflow below.
Configuration
| Variable |
Description |
{{DOCS_ROOT}} |
Path to resume master docs directory (e.g. resume-docs/master/) |
{{PROMPTS_ROOT}} |
Path to extraction/tailoring prompt templates (e.g. resume-docs/prompts/) |
{{INDEX_ROOT}} |
Path to processing index files (e.g. resume-docs/_index/) |
{{CONVERSATIONS_DIR}} |
Path to exported conversation archives (if using content extraction) |
1. Resume Review
Purpose: Formatting/design, content quality, or job-specific content decisions.
Steps:
- Identify scope from the user: full review, formatting-only, content-only, or review against a specific job description.
- Load context: The resume (or path) and, if job-specific, the job posting. Optionally load master docs to check traceability to evidence.
- Apply standards from the tailored resume prompt and score using the resume quality rubric (target 100):
- Structure: header, headline, executive summary, competencies, experience (reverse chronological), education.
- Bullets: Action + Scope + Method + Outcome; quantified where possible.
- Tone: confident, precise, outcomes-oriented; no vague claims without evidence.
- ATS: single-column, parseable; no meta text in canonical output (no Status/DONE/draft).
- Deliver: Concise feedback (formatting / content / job-fit) with specific edits or suggestions; optionally include rubric score (0-100). If job-specific, note keyword coverage and gaps.
Output format:
- Summary (2-3 lines)
- By section or theme: what works, what to change, optional rewrite snippets
- Optional: requirement-coverage note if a JD was provided
2. Job Evaluation -- Scores, Recommendation, and Resume Revisions
Purpose: Score any job (browser-found or user-supplied JD) for fit and interview likelihood; give a clear apply/not-apply recommendation; for strong fits, output specific resume revision recommendations.
Steps:
Load: The job description/posting (and optional company context). Load the universal job rubric and the master docs needed to map evidence.
Score: Apply Rubric A (Ideal Job Match 0-100) and Rubric B (Interview Odds 0-100). Use only documented evidence from master docs.
Recommend: Map scores (and any deal-breakers) to exactly one tier:
- Definite No Go | Nah | Meh | Good Fit | Freaking Apply Now
Why: Provide 2-4 sentences citing both scores and 1-2 concrete factors that drove the recommendation.
Resume revision recommendations: Only if recommendation is Good Fit or Freaking Apply Now, produce:
- Summary edits
- Core competencies adjustments
- Experience bullets by role
- Keywords to integrate
- Cover letter/application notes
Every item must be specific, actionable, and traceable to master docs.
Deliver: Scores -> Recommendation tier -> Why -> [If Good Fit+] Resume revision recommendations (copy-paste applicable).
3. Rubric Development and Resume Adaptation Evaluation
Purpose: Create or refine scoring rubrics; evaluate how well a resume (or adaptation) matches a role or meets criteria.
Steps:
- Clarify: New rubric (e.g. for a role type or org) vs. evaluating an existing resume/adaptation against current rubrics.
- For rubric development:
- Read existing rubrics for structure reference.
- Propose or refine criteria and weights; document in the same style.
- For adaptation evaluation:
- Load the resume and the target job. Score using Ideal Job Match and Interview Odds.
- Check resume against the suggested resume revision recommendations (if any) and requirement coverage.
- List strong alignment and risks/gaps; suggest improvements.
- Deliver: Rubric draft or evaluation summary with scores, strengths, risks, and concrete next steps.
4. Content / Data Set Development from Conversations
Purpose: Enrich master docs and bullet list from conversation archives not yet processed.
Steps:
- Check processed list:
{{INDEX_ROOT}}/processed_conversations.md -- which prompts have been run for which conversation IDs.
- Identify gaps: Which conversation files are not yet processed for the prompts relevant to the user's goal (employers, timeline, skills, achievements, portfolio, AI, roles, story, descriptions, cross-check).
- One run = one (or few) conversation files in context. Do not load the entire archive at once. Use the prompt chain:
- Prompts 1-9: extract employers, timeline, skills, achievements, portfolio, AI work, roles, story, bullet-ready content into master docs and update the processed list.
- Prompt 10: cross-check and append gaps/conflicts to
{{INDEX_ROOT}}/gaps-and-conflicts.md; update processed list.
- Rules: Only write under
{{DOCS_ROOT}}. After each run, update the processed list for the prompt(s) used. No fabrication; cite conversation_id and angle for additions.
- Deliver: Short summary of what was processed, what was added, and any conflicts or suggested follow-ups.
5. Trust and Quality Audit of the Resume Pipeline
Purpose: Let the user trust the pipeline and know where quality is enforced.
Steps:
- Map the chain: List the pipeline components:
- Data: Master docs, populated by extraction prompts and cross-check.
- Tailoring: Tailored resume prompt with scoring, requirement mapping, evidence mapping, bullet rewriting, keyword integration.
- Output: Markdown (canonical), ATS plain text, optional HTML/PDF.
- Quality checkpoints:
- No fabrication; evidence only from master docs or approved extraction.
- Scoring before writing; requirement extraction -> evidence mapping -> positioning; quality checklist before finalizing.
- Traceability: every claim in a resume should map to a master doc or a cited conversation.
- Optional audit: For a given tailored resume, verify:
- Bullets trace to master bullet list or other master docs
- Keywords align with stated job
- Dates/titles consistent with master timeline and employers/roles doc
- Deliver: Short "chain overview" plus where quality is enforced; if an audit was run, report traceability and any issues.
Suggested File Structure
| Purpose |
Path |
| Master docs (source of truth) |
{{DOCS_ROOT}}/00-master-timeline.md ... 08-*.md |
| Universal job rubric |
{{DOCS_ROOT}}/10-universal-job-rubric-and-apply-recommendation.md |
| Ranked jobs |
{{DOCS_ROOT}}/11-ranked-jobs-with-apply-recommendations.md |
| Resume quality rubric (0-100) |
{{DOCS_ROOT}}/09-resume-quality-rubric.md |
| Master bullet list |
{{DOCS_ROOT}}/07-master-bullet-list.md |
| Tailored resume prompt |
{{PROMPTS_ROOT}}/prompt-11-master-tailored-resume.md |
| Extraction prompts |
{{PROMPTS_ROOT}}/prompt-01-employers.md ... prompt-10-cross-check.md |
| Processed list |
{{INDEX_ROOT}}/processed_conversations.md |
| Gaps/conflicts |
{{INDEX_ROOT}}/gaps-and-conflicts.md |
Quality Rules
- No fabrication -- every claim must trace to a master doc or cited conversation
- Score before you write -- always evaluate fit before tailoring
- One conversation at a time -- don't load entire archives
- Update the index -- mark conversations as processed after each run
- Traceability -- every bullet should map to evidence
1---2name: career3description: Career management skill covering resume coaching (review, score, tailor, audit), job evaluation, rubric development, content extraction from conversations, and pipeline quality assurance.4metadata:5 source: https://skillmd.com/skills/joshuashepherd/career6---78# Career910Career management including resume coaching, job evaluation, and career pipeline tools.1112## When to Invoke1314User can say things like:1516- "Run a resume review on [file or role]"17- "Review my resume for [formatting / design / content / this job]"18- "Score this job" / "Evaluate this job posting" / "Should I apply to this role?"19- "Help with rubric development" or "Evaluate this resume adaptation"20- "Develop the content set from remaining convos" or "Pull more from conversations"21- "Audit the resume chain" or "Can I trust this pipeline?"2223Route to the right workflow below.2425## Configuration2627| Variable | Description |28|----------|-------------|29| `{{DOCS_ROOT}}` | Path to resume master docs directory (e.g. `resume-docs/master/`) |30| `{{PROMPTS_ROOT}}` | Path to extraction/tailoring prompt templates (e.g. `resume-docs/prompts/`) |31| `{{INDEX_ROOT}}` | Path to processing index files (e.g. `resume-docs/_index/`) |32| `{{CONVERSATIONS_DIR}}` | Path to exported conversation archives (if using content extraction) |3334---3536## 1. Resume Review3738**Purpose:** Formatting/design, content quality, or job-specific content decisions.3940**Steps:**41421. **Identify scope** from the user: full review, formatting-only, content-only, or review against a specific job description.432. **Load context:** The resume (or path) and, if job-specific, the job posting. Optionally load master docs to check traceability to evidence.443. **Apply standards** from the tailored resume prompt and **score using** the resume quality rubric (target 100):45 - Structure: header, headline, executive summary, competencies, experience (reverse chronological), education.46 - Bullets: Action + Scope + Method + Outcome; quantified where possible.47 - Tone: confident, precise, outcomes-oriented; no vague claims without evidence.48 - ATS: single-column, parseable; no meta text in canonical output (no Status/DONE/draft).494. **Deliver:** Concise feedback (formatting / content / job-fit) with specific edits or suggestions; optionally include rubric score (0-100). If job-specific, note keyword coverage and gaps.5051**Output format:**5253- Summary (2-3 lines)54- By section or theme: what works, what to change, optional rewrite snippets55- Optional: requirement-coverage note if a JD was provided5657---5859## 2. Job Evaluation -- Scores, Recommendation, and Resume Revisions6061**Purpose:** Score any job (browser-found or user-supplied JD) for fit and interview likelihood; give a clear apply/not-apply recommendation; for strong fits, output specific resume revision recommendations.6263**Steps:**64651. **Load:** The job description/posting (and optional company context). Load the universal job rubric and the master docs needed to map evidence.662. **Score:** Apply Rubric A (Ideal Job Match 0-100) and Rubric B (Interview Odds 0-100). Use only documented evidence from master docs.673. **Recommend:** Map scores (and any deal-breakers) to exactly one tier:68 - **Definite No Go** | **Nah** | **Meh** | **Good Fit** | **Freaking Apply Now**694. **Why:** Provide 2-4 sentences citing both scores and 1-2 concrete factors that drove the recommendation.705. **Resume revision recommendations:** **Only** if recommendation is **Good Fit** or **Freaking Apply Now**, produce:71 - Summary edits72 - Core competencies adjustments73 - Experience bullets by role74 - Keywords to integrate75 - Cover letter/application notes7677 Every item must be specific, actionable, and traceable to master docs.7879**Deliver:** Scores -> Recommendation tier -> Why -> [If Good Fit+] Resume revision recommendations (copy-paste applicable).8081---8283## 3. Rubric Development and Resume Adaptation Evaluation8485**Purpose:** Create or refine scoring rubrics; evaluate how well a resume (or adaptation) matches a role or meets criteria.8687**Steps:**88891. **Clarify:** New rubric (e.g. for a role type or org) vs. evaluating an existing resume/adaptation against current rubrics.902. **For rubric development:**91 - Read existing rubrics for structure reference.92 - Propose or refine criteria and weights; document in the same style.933. **For adaptation evaluation:**94 - Load the resume and the target job. Score using Ideal Job Match and Interview Odds.95 - Check resume against the suggested resume revision recommendations (if any) and requirement coverage.96 - List strong alignment and risks/gaps; suggest improvements.974. **Deliver:** Rubric draft or evaluation summary with scores, strengths, risks, and concrete next steps.9899---100101## 4. Content / Data Set Development from Conversations102103**Purpose:** Enrich master docs and bullet list from conversation archives not yet processed.104105**Steps:**1061071. **Check processed list:** `{{INDEX_ROOT}}/processed_conversations.md` -- which prompts have been run for which conversation IDs.1082. **Identify gaps:** Which conversation files are not yet processed for the prompts relevant to the user's goal (employers, timeline, skills, achievements, portfolio, AI, roles, story, descriptions, cross-check).1093. **One run = one (or few) conversation files in context.** Do not load the entire archive at once. Use the prompt chain:110 - Prompts 1-9: extract employers, timeline, skills, achievements, portfolio, AI work, roles, story, bullet-ready content into master docs and update the processed list.111 - Prompt 10: cross-check and append gaps/conflicts to `{{INDEX_ROOT}}/gaps-and-conflicts.md`; update processed list.1124. **Rules:** Only write under `{{DOCS_ROOT}}`. After each run, update the processed list for the prompt(s) used. No fabrication; cite conversation_id and angle for additions.1135. **Deliver:** Short summary of what was processed, what was added, and any conflicts or suggested follow-ups.114115---116117## 5. Trust and Quality Audit of the Resume Pipeline118119**Purpose:** Let the user trust the pipeline and know where quality is enforced.120121**Steps:**1221231. **Map the chain:** List the pipeline components:124 - **Data:** Master docs, populated by extraction prompts and cross-check.125 - **Tailoring:** Tailored resume prompt with scoring, requirement mapping, evidence mapping, bullet rewriting, keyword integration.126 - **Output:** Markdown (canonical), ATS plain text, optional HTML/PDF.1272. **Quality checkpoints:**128 - No fabrication; evidence only from master docs or approved extraction.129 - Scoring before writing; requirement extraction -> evidence mapping -> positioning; quality checklist before finalizing.130 - Traceability: every claim in a resume should map to a master doc or a cited conversation.1313. **Optional audit:** For a given tailored resume, verify:132 - Bullets trace to master bullet list or other master docs133 - Keywords align with stated job134 - Dates/titles consistent with master timeline and employers/roles doc1354. **Deliver:** Short "chain overview" plus where quality is enforced; if an audit was run, report traceability and any issues.136137---138139## Suggested File Structure140141| Purpose | Path |142|---------|------|143| Master docs (source of truth) | `{{DOCS_ROOT}}/00-master-timeline.md` ... `08-*.md` |144| Universal job rubric | `{{DOCS_ROOT}}/10-universal-job-rubric-and-apply-recommendation.md` |145| Ranked jobs | `{{DOCS_ROOT}}/11-ranked-jobs-with-apply-recommendations.md` |146| Resume quality rubric (0-100) | `{{DOCS_ROOT}}/09-resume-quality-rubric.md` |147| Master bullet list | `{{DOCS_ROOT}}/07-master-bullet-list.md` |148| Tailored resume prompt | `{{PROMPTS_ROOT}}/prompt-11-master-tailored-resume.md` |149| Extraction prompts | `{{PROMPTS_ROOT}}/prompt-01-employers.md` ... `prompt-10-cross-check.md` |150| Processed list | `{{INDEX_ROOT}}/processed_conversations.md` |151| Gaps/conflicts | `{{INDEX_ROOT}}/gaps-and-conflicts.md` |152153---154155## Quality Rules156157- **No fabrication** -- every claim must trace to a master doc or cited conversation158- **Score before you write** -- always evaluate fit before tailoring159- **One conversation at a time** -- don't load entire archives160- **Update the index** -- mark conversations as processed after each run161- **Traceability** -- every bullet should map to evidence