# The DiligenceWorks Professional Toolkit

**12 free AI skills for deal professionals, compliance officers, municipal administrators, and anyone who evaluates claims for a living.**

Built by [DiligenceWorks](https://diligenceworks.online) — adversarial AI for deal diligence.

## What's Inside

These skills extend Claude with specialized knowledge for professional workflows that no other AI toolkit covers. Each skill is a standalone `SKILL.md` file — drop it into Claude Code, Claude Desktop, or any compatible AI coding tool.

### Tier 1 — Headline Skills

| Skill | Who It's For | What It Does |
|-------|-------------|-------------|
| **Municipal Change Order Reviewer** | City engineers, contract administrators, auditors | Evaluates construction change orders against CCA 1-2021 / CCDC 2 procedures |
| **Tender Bid Evaluation Matrix** | Procurement officers, contracting authorities | Builds scoring matrices with OECD bid-rigging indicators and abnormally-low detection |
| **LC Document Compliance Checker** | Trade finance officers, documentary credit teams | Checks trade documents against UCP 600 letter of credit requirements |
| **Charter Party Term Extractor** | Ship brokers, charterers, P&I correspondents | Extracts and structures commercial terms from chartering agreements |
| **Litigation Funding Case Screener** | Litigation funders, disputes lawyers | Structures case assessment: merits, quantum, enforceability, budget, returns |

### Tier 2 — Domain-Depth Differentiators

| Skill | Who It's For | What It Does |
|-------|-------------|-------------|
| **Sanctions Screening Decision Writer** | Compliance officers, sanctions analysts | Documents screening decisions to the *Tonzip* evidentiary standard |
| **Immigration Case File Organizer** | Immigration consultants (RCIC), settlement agencies | Structures client files against IRCC program requirements |
| **Legal Invoice Audit Prep** | General counsel, legal ops, CFOs | Structures billing data for review — block billing, rate compliance, engagement terms |
| **DDQ Response Structurer** | Fund managers, LP relations, allocators | Organizes fund data against the ILPA Due Diligence Questionnaire framework |

### Tier 3 — Cross-Cutting Utilities

| Skill | Who It's For | What It Does |
|-------|-------------|-------------|
| **Multi-Jurisdiction Regulatory Checklist** | Cross-border deal teams, compliance | Generates transaction-specific regulatory checklists by jurisdiction |
| **Risk Register Builder** | Deal teams, project managers, auditors | Builds structured risk registers for transactions and projects |
| **Stakeholder Authority Map** | M&A teams, municipal administrators, board secretaries | Maps decision-makers, approval thresholds, and delegation chains |

## Installation

### Claude Code / Claude Desktop

```bash
# Clone the toolkit
git clone https://github.com/diligenceworks/professional-toolkit.git

# Install a single skill (global)
cp -r professional-toolkit/skills/municipal-change-order ~/.claude/skills/

# Or install all skills
cp -r professional-toolkit/skills/* ~/.claude/skills/
```

### Claude.ai (Projects)

Upload the `SKILL.md` file from any skill folder to your Claude Project's knowledge base.

### Other AI Tools (Cursor, Codex, Windsurf)

Each `SKILL.md` is a plain markdown file. Copy its contents into your tool's system prompt, `CLAUDE.md`, or equivalent configuration file.

## About DiligenceWorks

DiligenceWorks builds adversarial AI for deal diligence. Our platform treats every claim as a claim, not a fact — running automated verification against the evidence before any human sees a green light.

These skills help you **organize, structure, and produce** professional documents. For **adversarial verification** of the claims within those documents — whether a change order's pricing is supported, whether a borrower's representations hold up, whether a fund's track record checks out — that's what [DiligenceWorks](https://diligenceworks.online) does.

**Singapore** | **Dubai** | **London** | **Toronto**

## License

MIT — use these skills however you want. Attribution appreciated but not required.

## Contributing

Found a gap? Built an extension? We welcome pull requests — especially from practitioners who use these workflows daily and can improve the domain knowledge.