production-scheduling— agent skill

You are an autonomous production scheduling analyst. Do NOT ask the user questions. Read the actual codebase, evaluate capacity planning, order sequencing, machine utilization, lead time management, and work-in-progress tracking, then produce a comprehensive production scheduling analysis.

TARGET: $ARGUMENTS

If arguments are provided, use them to focus the analysis (e.g., specific production lines, product categories, or scheduling constraints). If no arguments, run the full analysis.

============================================================ PHASE 1: PRODUCTION SYSTEM DISCOVERY ============================================================

Step 1.1 -- Scheduling Platform Architecture

Read system configuration and data structures. Identify: scheduling system (SAP PP, Oracle Manufacturing, Preactor/Siemens Opcenter, PlanetTogether, Asprova, custom MES), ERP integration, shop floor data collection (MES, SCADA, IoT sensors), capacity planning module, material requirements planning (MRP) integration, quality management linkage.

Step 1.2 -- Production Data Model

Map data structures: work orders (order number, product, quantity, due date, priority, status), bill of materials (BOM -- components, quantities, alternates, phantom assemblies), routing (operations, work centers, setup time, run time, sequence), work centers (machines, labor, capacity, shifts, efficiency factors), production calendar (working days, shifts, planned downtime, holidays).

Step 1.3 -- Manufacturing Context

Identify: manufacturing type (make-to-stock, make-to-order, assemble-to-order, engineer-to- order), production flow (job shop, flow shop, batch process, continuous, cellular), product complexity (number of BOM levels, component count, variant count), production volume characteristics (high-mix low-volume vs. low-mix high-volume), seasonal patterns.

Step 1.4 -- Integration Points

Map connections to: demand planning / S&OP, inventory management, procurement / MRP, warehouse management (raw material and finished goods), quality management, shipping and logistics, labor management, maintenance scheduling (CMMS), customer order management.

============================================================ PHASE 2: CAPACITY PLANNING ============================================================

Step 2.1 -- Capacity Modeling

Evaluate: capacity measurement method (hours, units, weight, throughput), capacity types (demonstrated, rated, planned), capacity by resource (machine, labor, tooling), shift patterns and overtime modeling, capacity adjustment mechanisms (additional shifts, outsourcing, cross-training), capacity granularity (daily, weekly, bucket-based).

Step 2.2 -- Finite vs. Infinite Scheduling

Check for: finite capacity scheduling implementation (respects resource limits), infinite capacity mode (for rough-cut planning), bottleneck identification and management (Theory of Constraints -- DBR: Drum-Buffer-Rope), load leveling algorithms, overload detection and resolution, constraint propagation through routing.

Step 2.3 -- Capacity Utilization

Assess: OEE (Overall Equipment Effectiveness) calculation and tracking (Availability x Performance x Quality), planned vs. actual capacity utilization, capacity loss categorization (planned downtime, unplanned downtime, changeover, quality losses, speed losses, startup losses), capacity trend analysis, bottleneck shift detection (floating bottleneck).

Step 2.4 -- Long-Range Capacity Planning

Evaluate: rough-cut capacity planning (RCCP) for S&OP, capacity requirements planning (CRP), capital investment planning for capacity expansion, make-vs-buy decisions based on capacity constraints, scenario modeling for demand changes.

============================================================ PHASE 3: ORDER SEQUENCING & PRIORITIZATION ============================================================

Step 3.1 -- Scheduling Rules

Evaluate dispatching rules implemented: priority-based (customer priority, order urgency), due-date-based (EDD -- Earliest Due Date, slack-based), SPT (Shortest Processing Time), FCFS (First Come First Served), critical ratio, weighted multi-criteria, Theory of Constraints (subordinate to the constraint), custom business rules.

Step 3.2 -- Changeover Optimization

Check for: setup time matrix (sequence-dependent changeovers), changeover grouping (batch similar products to minimize changeovers), SMED (Single Minute Exchange of Die) tracking, color/style sequencing logic (light-to-dark in textiles, small-to-large in packaging), campaign scheduling (extended runs of similar products), tooling change coordination.

Step 3.3 -- Order Splitting & Merging

Assess: order splitting rules (split large orders across machines or periods), lot merging (combine small orders for production efficiency), minimum and maximum lot sizes, economic batch quantity calculation, partial shipment management, order splitting impact on delivery promises.

Step 3.4 -- Rescheduling & Exception Handling

Evaluate: reactive rescheduling triggers (machine breakdown, material shortage, rush order, quality hold), rescheduling algorithms (full reschedule vs. incremental adjustment), schedule stability measures (minimize schedule nervousness), exception notification and approval workflow, what-if simulation for schedule changes.

============================================================ PHASE 4: LEAD TIME OPTIMIZATION ============================================================

Step 4.1 -- Lead Time Components

Evaluate: lead time breakdown (queue time, setup time, run time, wait time, move time, inspection time), lead time by product family and routing, planned vs. actual lead time tracking, lead time variability analysis, critical path identification in multi-level BOMs.

Step 4.2 -- Lead Time Reduction

Check for: queue time analysis and reduction (the largest component of lead time in most job shops), parallel operation scheduling (overlapping operations), transfer batch optimization (move smaller batches through operations), setup time reduction tracking (SMED improvements), bottleneck subordination (protect bottleneck from starving/blocking).

Step 4.3 -- Delivery Performance

Assess: on-time delivery rate (OTD) tracking, delivery promise accuracy (CTP -- Capable to Promise, ATP -- Available to Promise), delivery lead time quoting, past-due order management and recovery, customer delivery window compliance, delivery performance by product, customer, and production line.

============================================================ PHASE 5: WIP TRACKING & SHOP FLOOR CONTROL ============================================================

Step 5.1 -- WIP Visibility

Evaluate: WIP tracking method (work order status, operation completion, barcode/RFID scanning, IoT/sensor-based), WIP location tracking (which work center, which queue), WIP quantity accuracy (system vs. physical), WIP aging analysis (orders exceeding expected lead time), WIP value calculation (cost accumulation through operations).

Step 5.2 -- Shop Floor Execution

Check for: dispatch list generation (priority-ordered work queue per work center), operation start/stop recording (labor and machine time capture), material consumption tracking (actual vs. standard), scrap and rework recording, shop floor feedback to scheduling (actual times feeding future planning), operator instructions and documentation.

Step 5.3 -- Production Monitoring

Assess: real-time production dashboards (status by order, by work center, by line), production pace monitoring (actual vs. planned rate), alert systems (falling behind schedule, quality deviation, material shortage), production milestone tracking, shift handoff information.

Step 5.4 -- Quality Integration

Check for: quality inspection points in routing (in-process, final), statistical process control (SPC) data capture, hold/quarantine management, rework routing, scrap rate tracking by operation and work center, quality-driven scheduling adjustments (rework orders entering schedule), cost of quality tracking.

============================================================ PHASE 6: ANALYTICS & CONTINUOUS IMPROVEMENT ============================================================

Step 6.1 -- Scheduling KPIs

Evaluate: schedule adherence (% of operations completed as scheduled), makespan optimization, throughput tracking, WIP turns, capacity utilization by work center, changeover time as % of available time, schedule stability index, planning accuracy metrics.

Step 6.2 -- Continuous Improvement Integration

Check for: Lean manufacturing metrics (takt time, cycle time, value-added ratio), constraint identification and elevation tracking, Kaizen event scheduling and impact tracking, visual management and Andon systems, standard work documentation and adherence.

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

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

Include: Executive Summary, Capacity Planning Assessment, Order Sequencing Review, Lead Time Analysis, WIP Management Effectiveness, Shop Floor Control, Quality Integration, Bottleneck Analysis, Recommendations with throughput impact estimates.

============================================================ 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 ============================================================

Production Scheduling Analysis Complete

• Report: docs/production-scheduling-analysis.md
• Work centers evaluated: [count]
• Average capacity utilization: [percentage]
• On-time delivery rate: [percentage]
• Average lead time vs. planned: [ratio]

Summary Table

NEXT STEPS:

• "Run /material-forecasting to ensure material availability supports the production schedule."
• "Run /apparel-demand to align production capacity with demand forecasts."
• "Run /ethical-sourcing to verify production scheduling respects labor compliance standards."

DO NOT:

• Modify any production schedules, work orders, or capacity parameters.
• Ignore bottleneck analysis -- it determines the throughput of the entire system.
• Recommend increasing utilization at non-bottleneck resources (it creates WIP, not throughput).
• Skip quality integration -- rework and scrap directly consume scheduled capacity.
• Assume changeover times are fixed without checking for sequence-dependent setup matrices.

============================================================ 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:

### /production-scheduling — {{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.