Client Reports — Professional Client Reporting for Consultants

Generate professional monthly retainer reports, KPI dashboards, quarterly business reviews, and variance analyses from CSV metrics data — as an alternative to AgencyAnalytics/Databox for consultants and small agencies.

Turn a 3-hour monthly reporting workflow into 15 minutes. included with your subscription. All data stays local.

Installation

1. Download the pf-client-reports.plugin file
2. Open Claude Desktop and navigate to Settings > Plugins
3. Click Install Plugin and select the downloaded .plugin file
4. The plugin will be installed and available immediately

Note: All data stays local on your machine. No external API calls or cloud storage required.

Quick Start

1. Initialize Workspace

/reports:setup "Your Firm Name"

This creates folders (inbound/, processing/, outbound/, .reports/) and default configuration.

2. Place Metrics CSV

Save your metrics to inbound/metrics.csv:

metric_name,current_value,previous_value,target,category
Website Visitors,12500,11200,15000,engagement
Revenue,125000,112000,150000,revenue
Conversion Rate,3.5,3.2,5.0,conversion

3. Generate Reports

/reports:full "Client Name" metrics.csv

This generates:

• Monthly Report — Executive summary, KPI tables, variance commentary, recommendations
• KPI Dashboard — One-page traffic-light status indicators
• Variance Analysis — Deep-dive into significant movements

Check outbound/ for generated DOCX files.

Commands

Command	Purpose
/reports:setup [firm-name]	Initialize workspace with folders, config, and dependencies
/reports:import <client-name> [csv]	Import CSV metrics, auto-categorize KPIs, calculate deltas
/reports:monthly <client-name> [period]	Generate detailed monthly retainer report DOCX
/reports:kpi <client-name> [period]	Generate one-page KPI dashboard with traffic-light status
/reports:qbr <client-name> [quarter]	Generate quarterly business review with 3-month trends
/reports:variance <client-name> [period]	Analyze significant metric movements with narratives
/reports:full <client-name> [csv]	Run full pipeline: import → monthly → KPI → variance
/reports:status	View workspace status, configured clients, generated reports

Features

Monthly Reports

• Professional DOCX formatting with firm branding
• Executive summary narrative (LLM-synthesized)
• KPI performance table with all metrics
• Variance commentary for top 5 movers
• Recommended next steps (3-5 data-driven actions)

KPI Dashboard

• One-page design for printing/sharing
• Traffic-light status indicators (green/yellow/red based on targets)
• Current vs target vs previous period values
• Top 3 movers highlights

Quarterly Business Reviews

• 3-month trend tables
• Goal attainment score (% of targets met)
• Strategic recommendations
• Next-quarter priorities

Variance Analysis

• Deep-dive into significant movements (>threshold %)
• Contributing factor hypotheses
• Anomaly detection (>50% changes, direction reversals)
• Waterfall-style narrative explanations

Workspace Management

• Status overview with client count, metric count, generated reports
• Pipeline health checks
• Audit logs (NDJSON format)

AI-Powered Features

• Pyramid Principle Executive Summaries: Generates answer-first narrative summaries using the Minto Pyramid Principle (1987) with SCQA framework (Situation-Complication-Question-Answer) — the standard McKinsey/BCG consulting narrative structure
• Statistical Process Control Variance Analysis: Identifies and explains significant metric movements using Shewhart control chart methodology (1931) with Western Electric Rules (1956) for anomaly detection and Ishikawa fishbone (1968) root cause categories adapted for digital metrics
• Balanced Scorecard KPI Taxonomy: Auto-categorizes metrics using Kaplan & Norton (1996) 4-perspective framework with 70+ domain keywords spanning GA4 dimensions, HubSpot lifecycle stages, and consulting engagement metrics
• Evidence-Based Traffic-Light Scoring: Ranks KPIs against targets using Locke & Latham Goal-Setting Theory (1990) thresholds: Green >=90% (on track), Yellow 70-89% (at risk), Red <70% (off track)
• Cohen's d Magnitude Classification: Classifies all metric movements using adapted effect size conventions (Cohen, 1988): small (<5%), medium (5-15%), large (>15%) for standardized significance language
• SMART Strategic Recommendations: Generates Specific, Measurable, Assignable, Realistic, Time-related recommendations (Doran, 1981) prioritized by Eisenhower Matrix (urgent/important quadrant analysis)
• Gainsight QBR Methodology: Quarterly reviews follow Success Plan structure with OKR-formatted next-quarter priorities (Doerr, 2018) and Nelson Rules (1984) trend detection
• Cluster Anomaly Detection: Flags systemic issues when 3+ metrics in the same Balanced Scorecard category all decline — distinguishing systemic problems from isolated fluctuations
• Pareto-Focused Analysis: Applies the vital few principle (Juran, 1951) — top movers analysis focuses on the 20% of metrics driving 80% of portfolio movement
• Tufte-Informed Dashboard Design: KPI dashboards follow data visualization best practices — maximum data-ink ratio, Cleveland & McGill perception hierarchy (1984) for encoding choices

Features Comparison

Feature	Client Reports	AgencyAnalytics	Databox	Klipfolio
Monthly retainer reports	✓ Full	✓ Full	✓ Full	✓ Full
KPI dashboard summary	✓ Full	✓ Full	✓ Full	✓ Full
Quarterly business reviews	✓ Full	△ Partial	✗ None	✗ None
Variance analysis	✓ Full	✗ None	△ Partial	△ Partial
Real-time dashboards	✗ None	✓ Full	✓ Full	✓ Full
Live data integrations	✗ None	✓ 80+	✓ 130+	✓ 130+
White-label branding	△ Partial	✓ Full	△ Partial	✓ Full
DOCX report export	✓ Full	△ Partial	△ Partial	✗ None
Data privacy (local only)	✓ Yes	✗ No	✗ No	✗ No
Customizable templates	✓ Full	△ Partial	△ Partial	✓ Full
No subscription required	✓ Yes	✗ No	✗ No	✗ No
Cost	included with your subscription	$948-6,000+/yr	$1,908-4,788/yr	$960-5,040/yr

Estimated Cost per Use

Disclaimer: Token estimates are approximate and based on typical usage patterns measured from skill prompt sizes. Actual costs vary with input data size, conversation length, and complexity. Estimates use Claude Sonnet 4.6 pricing ($3/1M input, $15/1M output). Cowork and Claude Desktop subscription users (Pro/Max/Team) are not charged per-token — these estimates apply only to direct Anthropic API usage. Running stages individually in fresh sessions uses fewer input tokens than running the full pipeline sequentially, because pipeline mode accumulates conversation history across stages.

Per skill (run individually in a fresh session):

Stage	Skill Prompt	User Input	Total Input	Output	Est. Cost
reports-kpi	~4.8K	~800	~9.3K	~6.0K	~$0.12
reports-import	~6.7K	~2.0K	~12.4K	~2.0K	~$0.07
reports-variance	~6.7K	~800	~11.6K	~6.0K	~$0.12
reports-qbr	~5.6K	~800	~10.5K	~6.0K	~$0.12
reports-monthly	~8.4K	~800	~12.8K	~6.0K	~$0.13
Standalone total			~56.5K	~26.0K	~$0.56

Full pipeline (all stages in one session — context accumulates):

Stage	Base Input	+ History	Total Input	Output	Est. Cost
reports-kpi	~9.7K	0	~9.7K	~6.0K	~$0.12
reports-import	~12.9K	~6.8K	~19.7K	~2.0K	~$0.09
reports-variance	~11.6K	~10.8K	~22.4K	~6.0K	~$0.16
reports-qbr	~10.5K	~17.6K	~28.1K	~6.0K	~$0.17
reports-monthly	~13.3K	~24.4K	~37.7K	~6.0K	~$0.20
Pipeline total			~117.6K	~26.0K	~$0.74

Running the full pipeline once typically costs $0.52–$0.97 in API tokens (Claude Sonnet 4.6).

Known Limitations

1. No Real-Time API Integrations

Impact: Cannot pull live data from Google Analytics, HubSpot, etc. Users must export CSVs manually.

Workaround: CSV is the universal export format. Most consultants already export to spreadsheets for analysis. See CSV template in /reports:help csv-format.

2. No Persistent Database / Cloud Sync

Impact: Metrics stored as local JSON files. No cloud backup or cross-device sync.

Workaround: Data persists between Cowork sessions in your workspace folder. Backup .reports/ directory manually or use Cowork's folder sync features.

3. No Automated Scheduling (Within Plugin)

Impact: Cannot auto-generate reports on a schedule like SaaS tools.

Workaround: Use Cowork's /schedule command:

/schedule '/reports:full "Client Name" metrics.csv' --cron '0 9 1 * *'

(Runs monthly at 9 AM on 1st of each month)

4. No White-Label Client Portals

Impact: Cannot provide clients with a branded login to view live dashboards.

Workaround: Generate branded DOCX reports and email as attachments or convert to PDF. Most SMB consultants deliver reports as files anyway — not live portals.

5. LLM Context Window (~115K Tokens)

Impact: Very large datasets (100+ KPIs across 12+ months) may need chunking.

Workaround: Import skill auto-chunks by KPI category. Individual reports cap at most-relevant metrics. Use /reports:status to monitor data volume.

Architecture

Pattern: Hub-and-spoke with independent report skills

CSV → [import] → metrics.json → [monthly | kpi | qbr | variance] → DOCX reports

• Import is the hub: Loads CSV, validates, calculates deltas deterministically
• Report skills are spokes: Each runs independently after import
• Users can run just /reports:monthly or /reports:qbr without full pipeline

Data Flow

1. User places CSV in inbound/ folder
2. Import skill reads CSV, validates, auto-categorizes KPIs, calculates deltas (Python, not LLM), saves JSON
3. Report skills load metrics JSON independently and generate DOCX documents
4. User retrieves DOCX from outbound/ folder and emails to client

All math (deltas, percentages, averages) is deterministic and reproducible. Only narrative text (summaries, recommendations, explanations) uses LLM synthesis.

Performance

• Token budget: ~45-60K tokens for full pipeline (monthly import + all 4 reports)
• Execution time: 30 seconds to 5 minutes depending on dataset size
• Storage: Typical workspace grows ~5-10 MB per client per year

Well within Cowork's 115K context limits — no risk of overflow on standard usage.

Context Efficiency

• Setup: ~3.5K tokens (one-time)
• Import: ~10K tokens per client per month
• Monthly report: ~13.5K tokens per generation
• KPI dashboard: ~8.5K tokens per generation
• QBR: ~14.5K tokens (3-month aggregation)
• Variance analysis: ~9.5K tokens per analysis

Total for full monthly workflow (import + all reports for one client): ~45K tokens

Support & Troubleshooting

Workspace Issues

"Workspace not initialized"

/reports:setup

Creates all required folders and config.

"python-docx required"

/reports:setup

Auto-installs python-docx dependency.

"CSV has no header row" Add header row to your CSV:

metric_name,current_value,previous_value,target

Data Issues

"No metrics found for client/period"

/reports:status

Check what periods are imported. Re-run /reports:import if needed.

"Non-numeric value" warning Update CSV to have numeric values. Use null or empty cell for missing values. Re-import with /reports:import.

"Duplicate metric names" Metrics are renamed to {name}_2, {name}_3. Update CSV to use unique metric names.

Report Generation

"DOCX file is invalid" Workspace may have disk space issues or corrupted output. Run /reports:status to check workspace health.

"Report looks wrong/incomplete" Check log files: .reports/logs/reports.ndjson. Review input CSV for data quality.

CSV Template

Download or use this CSV template:

metric_name,current_value,previous_value,target,category,period
Website Visitors,12500,11200,15000,engagement,2026-02
Unique Users,8900,8200,10000,engagement,2026-02
Page Views,45000,42000,50000,engagement,2026-02
Revenue,125000,112000,150000,revenue,2026-02
Cost Per Acquisition,45,48,40,conversion,2026-02
Conversion Rate,3.5,3.2,5.0,conversion,2026-02
Customer Lifetime Value,12500,12000,15000,retention,2026-02
Churn Rate,2.1,2.3,1.5,retention,2026-02

Rules:

• All metric_names must be unique
• current_value and previous_value are required (numeric)
• target and category are optional
• period can be omitted if in filename (e.g., february-2026-02.csv)

KPI Categories

By default, metrics are auto-categorized into:

• revenue: revenue, mrr, arr, sales, bookings, growth_rate
• engagement: visitors, sessions, pageviews, users, engagement_rate
• conversion: conversion, ctr, cpc, roas, cost_per_acquisition
• retention: churn, retention, lifetime_value, repeat_rate, satisfaction
• custom: (anything else)

Edit .reports/config.json to customize categories.

Customization

Firm Profile

Edit .reports/config.json:

"firm_profile": {
  "firm_name": "Your Consulting Firm",
  "firm_description": "Your tagline",
  "contact_email": "contact@yourfirm.com",
  "website": "https://yourfirm.com"
}

Report Defaults

"report_defaults": {
  "variance_threshold_percent": 10,
  "currency_symbol": "$",
  "include_recommendations": true,
  "max_metrics_per_page": 20
}

KPI Categories

"kpi_categories": {
  "revenue": {
    "label": "Revenue & Growth",
    "metrics": ["revenue", "mrr", "arr", ...],
    "color": "#2ecc71"
  }
}

Version

Client Reports v1.1

• Plugin Factory: v0.7.3
• Python requirement: 3.8+
• python-docx: 0.8.11+

Contributing

To contribute improvements or report bugs, use Cowork's built-in feedback system or open an issue on the plugin factory repository.

FAQ

Q: Can I use this with real-time data sources? A: Not directly. Export to CSV first, then import. Consider scheduling automated exports via your data source's API or scheduled reports feature.

Q: Can clients view reports without downloading DOCX? A: Not in this version. Reports are file-based, not web portals. Send DOCX via email or convert to PDF for web sharing.

Q: How do I add my company logo to reports? A: Edit config.json to reference a logo file. The plugin uses python-docx programmatic generation, so logo placement is flexible. See templates/README.md.

Q: Can I use this for multiple agencies/teams? A: Yes. Create separate workspace folders for each agency. Each has its own config and client registry.

Q: How far back does historical data go? A: As far as you import. QBRs require 3 months; monthly reports work with any single period. Plan to import data retrospectively if needed.

Q: What if I have 500+ metrics? A: Reports auto-chunk by category. See detailed analysis via /reports:variance. Monitor workspace size via /reports:status.

Support Resources

• Setup help: /reports:setup --help or see Setup Skill documentation
• CSV format: /reports:import --help or CSV Template section above
• Workspace status: /reports:status shows configured clients, periods, and pipeline health
• Logs: Check .reports/logs/ for detailed audit trail

---

Built with Plugin Factory | Designed for Cowork by Anthropic

Important Disclaimers

• AI-Generated Content: This plugin uses AI (LLM) technology which can produce inaccurate or incomplete outputs. All content should be treated as a starting point and reviewed for accuracy before use.
• Not Professional Advice: Outputs do not constitute legal, financial, tax, medical, or other professional advice. Consult qualified professionals before making decisions based on generated content.
• No Compliance Guarantee: References to industry standards, regulations, or guidelines are for informational purposes only. This plugin does not guarantee compliance with any law or regulation. Users are responsible for verifying all outputs meet their specific regulatory requirements.
• No Endorsement or Affiliation: Mention of third-party products, standards, or organizations does not imply endorsement, partnership, or certification by those entities.