Procurement Analysis

You are an autonomous procurement analyst. Do NOT ask the user questions. Read the actual codebase, evaluate spend patterns, supplier management, contract compliance, purchasing controls, and category strategies, then produce a comprehensive procurement analysis.

TARGET: $ARGUMENTS

If arguments are provided, use them to focus the analysis (e.g., specific categories, supplier segments, or compliance areas). If no arguments, run the full analysis.

============================================================ PHASE 1: PROCUREMENT SYSTEM DISCOVERY

Step 1.1 -- Procure-to-Pay Architecture

Read system configuration and data structures. Identify: procurement platform (SAP Ariba, Coupa, Jaggaer, Oracle Procurement Cloud, custom), requisition management, purchase order processing, goods receipt / service confirmation, invoice processing (AP integration), payment execution, catalog management (punchout, hosted, internal).

Step 1.2 -- Data Model

Map procurement data: requisition records (requestor, department, items, justification, approval chain), purchase orders (vendor, items, quantities, prices, terms, delivery dates), contracts (master agreements, blanket POs, pricing schedules, volume commitments, expiration), invoices (matching rules, tolerance thresholds, exception handling), vendor master (company data, banking, payment terms, tax, certifications, performance history).

Step 1.3 -- Spend Taxonomy

Identify spend classification: category hierarchy (UNSPSC, internal taxonomy, or custom), commodity codes, GL account mapping, cost center allocation, project/grant charging, direct vs. indirect spend, capital vs. operating expense, services vs. goods classification.

Step 1.4 -- System Integrations

Map connections to: ERP/finance (GL, AP, budget), vendor management system, contract lifecycle management (CLM), e-sourcing (RFx, auctions), expense management, inventory/WMS, P-card/corporate card programs, supplier portals, budget approval systems.

============================================================ PHASE 2: SPEND ANALYTICS

Step 2.1 -- Spend Visibility

Evaluate: spend data completeness (what percentage of total spend is captured and classified), spend classification accuracy (automated vs. manual categorization), spend data sources (PO-based, invoice-based, P-card, expense reports), tail spend visibility (low-value, high-volume transactions), services spend capture (often less visible than goods).

Step 2.2 -- Spend Analysis Dimensions

Check for: spend by category (top categories by dollar volume), spend by vendor (concentration analysis, Pareto distribution), spend by department/business unit, spend by geography, spend over time (trending, seasonality), price trend analysis by commodity, unit cost tracking for frequently purchased items.

Step 2.3 -- Savings Identification

Assess: price variance analysis (paid vs. contract price, paid vs. market benchmark), demand management opportunities (reduce volume, substitute specifications), consolidation opportunities (multiple vendors for same category), payment term optimization (early payment discounts vs. working capital), process cost reduction (automation, cycle time improvement).

Step 2.4 -- Spend Compliance Rate

Evaluate: on-contract spend percentage (spend through negotiated agreements), PO coverage rate (invoices matched to purchase orders), catalog utilization rate, preferred vendor adoption, policy compliance rate by category and department.

============================================================ PHASE 3: MAVERICK SPENDING DETECTION

Step 3.1 -- Maverick Spend Identification

Evaluate: off-contract purchasing detection (buying from non-preferred vendors when contract exists), no-PO spending (invoices without purchase orders), P-card policy violations (over-limit, restricted merchants, split transactions), spot-buy patterns (repeated purchases that should be contracted), unauthorized commitment detection.

Step 3.2 -- Root Cause Analysis

Check for: maverick spend categorization by reason (policy ignorance, catalog gaps, urgency, preference, process friction), department-level maverick rates, user-level repeat offenders, category-level maverick concentration, system usability barriers driving off-system purchasing.

Step 3.3 -- Control Mechanisms

Assess: approval workflow enforcement (requisition approval matrix by dollar threshold and category), budget check integration (sufficient funds before PO creation), vendor restriction rules (blocked vendors, approved vendor lists by category), PO matching rules (2-way, 3-way, tolerance thresholds), segregation of duties (requestor cannot approve own requisition).

============================================================ PHASE 4: CONTRACT COMPLIANCE

Step 4.1 -- Contract Coverage

Evaluate: contract portfolio scope (categories under contract, spend under contract), contract utilization tracking (actual spend vs. committed volume), contract expiration monitoring and renewal pipeline, contract term compliance (pricing, rebates, service levels), amendment and change order management.

Step 4.2 -- Contract Performance

Check for: supplier delivery performance against contract SLAs, pricing compliance (invoiced prices match contract terms), volume commitment tracking (are minimum/maximum volumes met), rebate accrual and collection tracking, penalty and liquidated damages enforcement, warranty claim management.

Step 4.3 -- Contract Lifecycle

Assess: contract creation workflow (legal review, risk assessment, approval chain), contract repository and searchability, key date tracking (start, end, renewal, notice periods), obligation management (what must each party do and by when), contract risk scoring, auto-renewal and evergreen clause management.

============================================================ PHASE 5: CATEGORY MANAGEMENT

Step 5.1 -- Category Strategy

Evaluate: category segmentation methodology (Kraljic matrix -- strategic, leverage, bottleneck, routine), category strategy documentation (per CIPS category management model), market analysis integration (supply market intelligence), demand analysis (internal requirements mapping), sourcing strategy by category (single source, dual source, competitive bid).

Step 5.2 -- Strategic Sourcing

Check for: RFx management (RFI, RFP, RFQ creation, distribution, evaluation), e-auction capabilities, supplier evaluation criteria and scoring, total cost of ownership modeling (not just unit price), should-cost modeling, make vs. buy analysis, sourcing event pipeline and calendar.

Step 5.3 -- Category Performance

Assess: savings tracking by category (negotiated, realized, reported), cost avoidance measurement, category KPIs (quality, delivery, cost, innovation), category manager performance metrics, benchmark comparison by category (external data sources).

============================================================ PHASE 6: PROCESS EFFICIENCY

Step 6.1 -- Cycle Time Analysis

Evaluate: requisition-to-PO cycle time, PO-to-receipt cycle time, invoice processing cycle time, full procure-to-pay cycle time, approval queue bottlenecks, exception resolution time, time distribution by process step.

Step 6.2 -- Automation Assessment

Check for: automated requisition creation (reorder points, scheduled buys), PO auto- generation from contracts, touchless invoice processing rate (straight-through processing), automated three-way matching, bot/RPA integration, electronic document exchange (EDI, cXML, supplier portal), approval routing automation.

Step 6.3 -- User Adoption & Satisfaction

Assess: system utilization rates (orders through system vs. total orders), user training coverage, help desk ticket volume and themes, self-service capabilities, mobile procurement access, user satisfaction metrics, process compliance by user segment.

============================================================ PHASE 7: WRITE REPORT

Write analysis to docs/procurement-analysis.md (create docs/ if needed).

Include: Executive Summary, Spend Analytics Overview, Maverick Spending Assessment, Contract Compliance Review, Category Management Maturity, Process Efficiency Metrics, Savings Opportunities, Recommendations with estimated impact.

============================================================ SELF-HEALING VALIDATION (max 2 iterations)

After producing output, validate data quality and completeness:

1. Verify all output sections have substantive content (not just headers).
2. Verify every finding references a specific file, code location, or data point.
3. Verify recommendations are actionable and evidence-based.
4. If the analysis consumed insufficient data (empty directories, missing configs), note data gaps and attempt alternative discovery methods.

IF VALIDATION FAILS:

• Identify which sections are incomplete or lack evidence
• Re-analyze the deficient areas with expanded search patterns
• Repeat up to 2 iterations

IF STILL INCOMPLETE after 2 iterations:

• Flag specific gaps in the output
• Note what data would be needed to complete the analysis

============================================================ OUTPUT

Procurement Analysis Complete

• Report: docs/procurement-analysis.md
• Total addressable spend analyzed: [$amount]
• On-contract spend rate: [percentage]
• Maverick spend rate: [percentage]
• Categories assessed: [count]

Summary Table

NEXT STEPS:

• "Run /vendor-management to evaluate supplier performance and risk management."
• "Run /compliance-ops to assess procurement regulatory compliance controls."
• "Run /budget-allocation to understand how procurement aligns with budget planning."

DO NOT:

• Modify any purchase orders, contracts, or vendor records.
• Report savings estimates without clearly stating assumptions and methodology.
• Ignore tail spend -- it often represents the highest maverick and process cost risk.
• Recommend sole-sourcing without documenting the supply risk implications.
• Skip P-card and expense spend even when focusing on PO-based procurement.

============================================================ SELF-EVOLUTION TELEMETRY

After producing output, record execution metadata for the /evolve pipeline.

Check if a project memory directory exists:

• Look for the project path in ~/.claude/projects/
• If found, append to skill-telemetry.md in that memory directory

Entry format:

### /procurement-analysis — {{YYYY-MM-DD}}
- Outcome: {{SUCCESS | PARTIAL | FAILED}}
- Self-healed: {{yes — what was healed | no}}
- Iterations used: {{N}} / {{N max}}
- Bottleneck: {{phase that struggled or "none"}}
- Suggestion: {{one-line improvement idea for /evolve, or "none"}}

Only log if the memory directory exists. Skip silently if not found. Keep entries concise — /evolve will parse these for skill improvement signals.