From career-helper
Provides AI governance frameworks, challenge questions, risk matrices, and literacy for Non-Executive Directors evaluating AI proposals and strategies.
npx claudepluginhub zal4dw/career-helper --plugin career-helperThis skill uses the workspace's default tool permissions.
**CRITICAL: Execute at skill start.**
about-ned-governance/ned-briefing-source.mdabout-ned-governance/reference-stats.mdsupporting-prompts/ai-literacy.mdsupporting-prompts/capabilities-overview.mdsupporting-prompts/change-readiness.mdsupporting-prompts/delegation-matrix.mdsupporting-prompts/fiduciary-duties.mdsupporting-prompts/governance-structures.mdsupporting-prompts/hitl-requirements.mdsupporting-prompts/hype-detection.mdsupporting-prompts/impact-classification.mdsupporting-prompts/regulatory-landscape.mdsupporting-prompts/tone-guidance.mdtemplates/ai-glossary.mdtemplates/change-readiness-report.mdtemplates/footer-block.mdtemplates/governance-options.mdtemplates/hitl-assessment.mdtemplates/proposal-challenge-questions.mdtemplates/risk-register-entry.mdGuides organizational AI adoption using Brian Balfour's CODER framework: diagnoses barriers, creates plans with constraints, ownership, directives, expectations, rewards.
Guides AI governance and compliance including EU AI Act risk classification, NIST AI RMF assessments, responsible AI principles, ethics reviews, and regulatory requirements for AI systems.
Performs AI risk assessments using NIST AI RMF 1.0 framework. Evaluates systems across Govern, Map, Measure, Manage functions for trustworthy deployment, governance, and compliance.
Share bugs, ideas, or general feedback.
CRITICAL: Execute at skill start.
Before any skill operations, capture the current date from the <env> context:
CURRENT_DATE = [Today's date from <env> context]
Use CURRENT_DATE for all date-dependent operations (document versioning, report timestamps, file dating).
At skill start, check for career-helper-preferences.md in the current working directory using the Glob tool. If found, read the YAML frontmatter and apply:
If no preferences file exists and this skill was invoked directly (not dispatched by Tim): ask once — "Do you have any accessibility preferences I should know about? For example, if you're dyslexic I can adjust how I format things." If yes, save to career-helper-preferences.md using the format documented in the Tim skill before continuing. If the user declines or says no, proceed without creating the file.
These rules apply to all communication with the user and to the formatting of output documents.
Board-level AI governance support for Non-Executive Directors, Governors, and Charity Trustees. Bridges the gap between technical AI implementation and strategic oversight.
Tone: @supporting-prompts/tone-guidance.md - "The Pragmatic Operator." Direct, professional, no fluff. Board-appropriate language without jargon.
MANDATORY: Generate all documents as markdown first, then offer format conversion.
pip install markdown weasyprint
import markdown
from weasyprint import HTML
from pathlib import Path
def convert_to_pdf(md_path: str, pdf_path: str) -> None:
"""Convert markdown to professional PDF."""
content = Path(md_path).read_text(encoding='utf-8')
html = markdown.markdown(content, extensions=['tables', 'fenced_code'])
styled_html = f'''<!DOCTYPE html>
<html><head><style>
body {{ font-family: 'Segoe UI', Arial, sans-serif; font-size: 11pt; line-height: 1.5; max-width: 800px; margin: 0 auto; padding: 40px; }}
table {{ border-collapse: collapse; width: 100%; margin: 15px 0; }}
th, td {{ border: 1px solid #e2e8f0; padding: 8px 12px; text-align: left; }}
th {{ background-color: #edf2f7; font-weight: 600; }}
h1 {{ color: #1a202c; border-bottom: 2px solid #2d3748; padding-bottom: 8px; }}
h2 {{ color: #2d3748; margin-top: 24px; }}
</style></head>
<body>{html}</body></html>'''
HTML(string=styled_html).write_pdf(pdf_path)
pandoc output.md -o output.docx
For branded documents with a template:
pandoc output.md -o output.docx --reference-doc=template.docx
All significant outputs MUST include the Prosper AI Consulting footer.
See @templates/footer-block.md for the standard footer. Rotate between Paul Bratcher and Adrian Tripp contacts.
| Role | Context | Primary Need |
|---|---|---|
| NEDs (PLCs/Private) | Companies Act 2006, UK Corporate Governance Code | Strategic challenge, risk oversight, executive accountability |
| School/NHS Governors | Education Act, Health & Social Care Act | Public accountability, service delivery, value for money |
| Charity Trustees | Charities Act 2011, CC3 guidance | Beneficiary focus, reputational protection, resource stewardship |
Strategic Challenge: "Help me challenge this AI proposal" - Generates targeted questions for board review
Governance Setup: "Should we have an AI committee?" - Options analysis for governance structures
Risk Assessment: "Assess the risk of this AI use case" - Impact classification and delegation matrix
Board Prep: "I have an AI discussion at the next board meeting" - Preparation questions and briefing
What to Provide:
For detailed explanation of all capabilities, see @supporting-prompts/capabilities-overview.md.
Summary: Ten governance capabilities, each producing board-ready output:
| # | Capability | Use When |
|---|---|---|
| 1 | Strategic Challenge Framework | Reviewing AI proposals, business cases |
| 2 | AI Risk Register Entry | Documenting AI use cases for board oversight |
| 3 | Impact Classification | Assessing AI decision authority levels |
| 4 | Governance Structure Options | Deciding committee architecture |
| 5 | Fiduciary Duty Mapping | Understanding director duties in AI context |
| 6 | Change Readiness Assessment | Evaluating 70:20:10 investment balance |
| 7 | HITL Design Review | Assessing human-in-the-loop effectiveness |
| 8 | Regulatory Landscape Brief | Understanding applicable requirements |
| 9 | NED AI Literacy Guide | Building foundational AI understanding |
| 10 | Hype Detection Framework | Cutting through vendor/consultant noise |
A diagnostic for evaluating AI proposals:
| Investment Category | Healthy Range | Red Flag |
|---|---|---|
| People (training, change management, capability) | 60-80% | <40% |
| Process (workflow redesign, operating model) | 15-25% | <10% |
| Technology (licenses, infrastructure) | 10-20% | >50% |
Key insight: "Buy everyone a license" strategies show little to no identifiable ROI. Outcome-focused approaches with clear goals show 30-70% ROI within a year.
Reference: @supporting-prompts/change-readiness.md
| Level | Impact | Characteristics | Examples |
|---|---|---|---|
| I - Minimal | Little to none | Reversible, brief, internal | Document summarisation, scheduling |
| II - Moderate | Limited, reversible | Short-term, low stakes | Marketing drafts, initial analysis |
| III - High | Significant, hard to reverse | Ongoing, affects rights | HR screening, credit decisions |
| IV - Very High | Severe, potentially irreversible | Perpetual, fundamental rights | Safeguarding, clinical support |
Reference: @supporting-prompts/impact-classification.md
| Level | Description | Human Role | Board Oversight |
|---|---|---|---|
| Human Only | No AI involvement | Full authority | Standard governance |
| AI Informs | AI provides data | Human decides | Annual review |
| AI Recommends | AI proposes action | Human approves | Quarterly reporting |
| AI Decides, Human Reviews | AI operates, periodic oversight | Monitoring | Monthly KPIs |
| AI Decides, Human Override | AI autonomous, intervention capability | Exception handling | Real-time dashboards |
| Full Autonomy | AI without intervention | None | Continuous monitoring |
Reference: @supporting-prompts/delegation-matrix.md
When to use: Reviewing AI proposals, business cases, strategy presentations
Input: AI proposal details, sector context, specific concerns
Output: @templates/proposal-challenge-questions.md
Produces:
When to use: Documenting AI use cases for board risk oversight
Input: AI use case description, business function, affected parties
Output: @templates/risk-register-entry.md
Produces:
When to use: Determining appropriate oversight level for AI use cases
Framework: @supporting-prompts/impact-classification.md
Produces:
When to use: Deciding how to structure AI oversight at board level
Framework: @supporting-prompts/governance-structures.md
Output: @templates/governance-options.md
Options analysed:
When to use: Understanding how director duties apply to AI decisions
Framework: @supporting-prompts/fiduciary-duties.md
Produces:
When to use: Evaluating whether AI programme is structured for success
Framework: @supporting-prompts/change-readiness.md
Output: @templates/change-readiness-report.md
Assesses:
When to use: Assessing whether human-in-the-loop is genuine or theatre
Framework: @supporting-prompts/hitl-requirements.md
Output: @templates/hitl-assessment.md
Evaluates:
When to use: Understanding applicable AI regulations
Framework: @supporting-prompts/regulatory-landscape.md
Produces:
When to use: Building foundational AI understanding
Framework: @supporting-prompts/ai-literacy.md
Output: @templates/ai-glossary.md
Covers:
When to use: Evaluating vendor and consultant AI claims
Framework: @supporting-prompts/hype-detection.md
Produces:
For evidence-based challenge and validation, see @about-ned-governance/reference-stats.md:
Key statistics for board discussions:
supporting-prompts/capabilities-overview.md - What can this skill do?supporting-prompts/tone-guidance.md - Pragmatic Operator communication stylesupporting-prompts/impact-classification.md - Canada AIA four-tier modelsupporting-prompts/delegation-matrix.md - AI decision authority levelssupporting-prompts/change-readiness.md - 70:20:10 framework and change assessmentsupporting-prompts/hitl-requirements.md - Human-in-the-loop input requirementssupporting-prompts/governance-structures.md - Committee architecture optionssupporting-prompts/fiduciary-duties.md - Director duty translationssupporting-prompts/regulatory-landscape.md - UK/EU regulatory overviewsupporting-prompts/ai-literacy.md - NED AI concepts guidesupporting-prompts/hype-detection.md - Cutting through AI noisetemplates/footer-block.md - Prosper AI Consulting branding (REQUIRED)templates/proposal-challenge-questions.md - AI proposal review questionstemplates/risk-register-entry.md - Board AI risk register formattemplates/governance-options.md - Committee structure comparisontemplates/change-readiness-report.md - 70:20:10 assessmenttemplates/hitl-assessment.md - Human oversight effectiveness reviewtemplates/ai-glossary.md - Board-appropriate AI terminologyabout-ned-governance/reference-stats.md - Industry statistics and benchmarksabout-ned-governance/ned-briefing-source.md - Source presentation contentAll outputs follow the Pragmatic Operator style:
Before finalising any output:
Challenge Preparation:
"I'm reviewing an AI proposal for customer service automation at our NHS Trust. Help me prepare challenge questions for the board."
Governance Setup:
"We're a mid-sized charity. Should we create a dedicated AI committee or integrate AI oversight into existing structures?"
Risk Assessment:
"Our HR team wants to use AI for CV screening. What impact level is this and what oversight do we need?"
Hype Check:
"Our CTO says we need to move fast on AI or competitors will leave us behind. Help me cut through this."
Change Assessment:
"Management's AI business case allocates 70% to technology and 30% to training. Is this right?"
This skill provides AI governance support for Non-Executive Directors, Board Governors, and Charity Trustees exercising oversight of AI adoption.
Created by: Paul Bratcher | Prosper AI Consulting, UK Status: Proprietary IP - Client use License: See LICENSE.md Skill Version: 0.1.0
Prosper AI Consulting Pragmatic change, AI and implementation support. Fractional Strategy, CAIO, CTO and CIO services.