Completion Design Reviewer

Completion Design Reviewer

What this skill does

Analyzes completion design documents, frac stage sheets, and pump schedules to validate design consistency, flag deviations from engineering intent, compare against offset wells, and summarize stimulation parameters. Covers plug-and-perf completions, sliding sleeve systems, and multi-lateral configurations.

When to use

• User uploads completion design spreadsheets, frac programs, or pump schedules
• User wants to review or QC a completion design before execution
• User asks to summarize frac parameters for a well or pad
• User needs to compare completion design against offset performance
• User is conducting a post-frac analysis against the design

Trigger phrases

• "Review this completion design"
• "Check the frac stage parameters"
• "Summarize the pump schedule for this well"
• "Does this completion design match the offset wells?"
• "QC the perforation intervals on this stage sheet"
• "Compare planned vs. actual frac parameters"
• "Analyze proppant loading and fluid design for this pad"
• "Flag any completion design inconsistencies"

Inputs

• Completion design spreadsheets (XLSX, CSV)
• Frac stage sheets (PDF, XLSX)
• Pump schedule files
• Perforation design documents
• Optional: offset completion summary for benchmarking
• Optional: casing/tubing design for compatibility check

Outputs

• Stage-by-stage parameter table
• Design summary statistics (totals, averages, min/max per parameter)
• Inconsistency flags (stage spacing outliers, cluster count mismatches)
• Deviation warnings (parameters outside design tolerance)
• Offset comparison table if reference data provided
• Post-frac vs. design comparison if actual data provided
• Structured JSON output with source references

Parameters extracted and validated

Stage design

| Parameter | Expected range | Flag condition | |---|---|---| | Stage count | 1–100 stages | Flag if total 80 for typical laterals | | Stage length (ft) | 150–400 ft | Flag if 500 ft | | Cluster count per stage | 3–6 clusters | Flag if 8 | | Cluster spacing (ft) | 25–75 ft | Flag if 100 ft | | Perforation interval (ft) | 1–3 ft | Flag if > 5 ft per cluster |

Fluid design

| Parameter | Notes | |---|---| | Fluid type | Slickwater, hybrid, crosslinked, foam | | Fluid volume per stage (bbls) | Flag if 5,000 for typical lateral | | Total fluid volume (bbls) | Sum across all stages | | Fluid rate (bpm) | Flag if > 150 bpm (pump limitation) | | Acid pre-pad volume (gals) | For carbonate stimulation |

Proppant design

| Parameter | Notes | |---|---| | Proppant type | 100 mesh, 40/70, 30/50, resin-coated, ceramic | | Proppant concentration (PPA) | Flag if > 4 PPA max for slickwater | | Proppant per stage (lbs) | Typical: 200,000–2,000,000 lbs | | Total proppant (lbs) | Sum across all stages | | Proppant loading (lbs/ft) | Typical range: 500–3,000 lbs/ft of lateral |

Pressure and rate

| Parameter | Notes | |---|---| | ISIP (psi) | Flag if varies > 20% across stages | | Breakdown pressure (psi) | Flag if > max allowable treating pressure | | Max treating pressure (psi) | Flag if approaching wellbore limits | | Pump rate schedule | Validate ramp-up and ramp-down steps |

Mechanical design

| Parameter | Notes | |---|---| | Plug type and depth | Per-stage plug setting depths | | Plug to perf spacing (ft) | Flag if 25% from median

1. Cluster count consistency: Flag stages with cluster count different from design
2. Fluid volume ramp: Expected gradual increase toe-to-heel for toe-biased design
3. Pressure limits: Max treating pressure must not exceed casing burst / 1.25
4. ISIP trend: Should be relatively consistent (+/- 15%) without outliers unless geological reason
5. Proppant concentration: Slurry concentration must not exceed pump limits
6. Total volume check: Sum of per-stage volumes must equal total volume within 2%
7. Plug depth sequence: Plugs must be in ascending depth order (heel to toe)
8. Perf cluster count vs. design: Actual cluster count must match design count
9. Missing stages: Gap detection if stage numbering is non-sequential

Output schema

{
  "design_summary": {
    "well_name": "string",
    "total_stages": 0,
    "total_fluid_bbls": 0.0,
    "total_proppant_lbs": 0.0,
    "avg_fluid_per_stage_bbls": 0.0,
    "avg_proppant_per_stage_lbs": 0.0,
    "avg_stage_length_ft": 0.0,
    "avg_cluster_spacing_ft": 0.0,
    "proppant_loading_lbs_ft": 0.0
  },
  "stages": [
    {
      "stage_number": 0,
      "top_md_ft": 0.0,
      "base_md_ft": 0.0,
      "stage_length_ft": 0.0,
      "cluster_count": 0,
      "fluid_volume_bbls": 0.0,
      "proppant_lbs": 0.0,
      "max_rate_bpm": 0.0,
      "isip_psi": 0.0,
      "plug_depth_ft": 0.0
    }
  ],
  "validation_results": [
    {
      "check": "string",
      "status": "pass | warn | fail",
      "detail": "string",
      "affected_stages": [0]
    }
  ],
  "offset_comparison": {
    "offset_well": "string",
    "this_well_proppant_lbs_ft": 0.0,
    "offset_proppant_lbs_ft": 0.0,
    "this_well_fluid_bbls_stage": 0.0,
    "offset_fluid_bbls_stage": 0.0
  },
  "quality_flags": ["string"]
}

Example response

COMPLETION DESIGN REVIEW — Eagle Ford A12H
══════════════════════════════════════════
Total stages:       32
Total fluid:        72,000 bbls (avg 2,250 bbls/stage)
Total proppant:     38.4 MM lbs (1,200 lbs/ft lateral)
Avg cluster spacing: 45 ft  |  Avg stage length: 200 ft

VALIDATION RESULTS
───────────────────
✓ Stage spacing consistent (median 200 ft, all within 10%)
✓ Cluster count consistent (4 per stage, all 32 stages)
✓ Proppant concentration within limits (max 2.5 PPA)
⚠ Stages 1–3 (toe): fluid volume 18% below pad average — verify toe-bias intent
⚠ Stage 18: ISIP 5,840 psi vs. avg 6,850 psi — potential natural fracture intersection
✗ Stage 27: plug depth 11,421 ft overlaps perf interval 11,380–11,440 ft — REVIEW REQUIRED

OFFSET COMPARISON (vs. 3 offset wells)
──────────────────────────────────────
Proppant loading: 1,200 lbs/ft (offset avg: 1,050 lbs/ft) +14%
Fluid per stage:  2,250 bbls   (offset avg: 2,100 bbls)   +7%
Stage count:      32           (offset avg: 29)            +10%

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: ttracx
• Source: ttracx/oil-and-gas-claude-skills
• License: MIT

Install and usage instructions live in the source repository linked above.