agentleFS
florianbonnet14florianbonnet14/ThePowerOfAnalytics_ClaudeSkills/analytics-orchestrator/SKILL.md

analytics-orchestrator

Skill28 starsChanged 9 months agoMITView on GitHub

Intelligent entry point to the analytics ecosystem based on "The Power of Analytics" book by Florian Bonnet. Guides users to the right analytics agent based on their needs, recommends workflow sequences, and provides clear instructions for launching agents. Use when users need help choosing between analytics agents, want to start an analytics project but aren't sure where to begin, or ask "what can you help me with" in an analytics context. Also use when users mention analytics tasks like defining metrics, planning analysis, collecting results, creating presentations, or writing executive summaries.

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 8 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When to use: After defining a North Star metric, need to identify what drives the metric, preparing for root cause analysis, setting up …

    line 45
  2. Distinctive:

    When the orchestrator recommends an agent, it provides specific instructions:

    line 120
  3. Distinctive:

    If unclear: Ask 1-2 discovery questions to understand stage and goal.

    line 153
  4. Distinctive:

    Explain the workflow: if multiple agents needed

    line 161
  5. Distinctive:

    One or two questions maximum: Don't overwhelm with discovery questions

    line 187
  6. Distinctive:

    Clear next steps: Always end with actionable instructions

    line 188
  7. Distinctive:

    All agents assume basic analytics literacy (no beginner tutorials included)

    line 199
  8. Distinctive:

    Don't use the orchestrator for:

    line 203

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Analytics Orchestrator

Welcome to the analytics ecosystem based on "The Power of Analytics" by Florian Bonnet. This orchestrator helps you navigate seven specialized analytics agents and find the right workflow for your needs.

How This Works

Rather than choosing agents yourself, describe what you're trying to accomplish. This orchestrator will:

  1. Understand your analytics challenge
  2. Recommend the right agent(s) and sequence
  3. Explain why each agent is recommended
  4. Provide clear instructions to launch each agent
  5. Free text entry for the user to write what they want to achieve in their own words

Discovery Questions

To route you effectively, the orchestrator starts by asking what is the users goal using the askquestion tool with options:

  1. Find or update my north star metric
  2. Develop my KPI Tree
  3. Plan an analysis
  4. Summarize my analysis and build a presentation
  • What stage are you at? (Planning, analyzing data, communicating results, or something else?)
  • What's your goal? (Define metrics, structure analysis, create presentation, etc.)

Based on your answers, the orchestrator launches the appropriate agents and workflow (do not ask confirmation from the user)

Available Agents

Strategic Foundation Agents

North Star Metric Advisor (north-star-metric-advisor)

  • What it does: Helps identify and validate the single most important metric that captures customer value delivery and predicts business success
  • When to use: Starting a new team/product, existing product lacks clear success metric, team needs to define what success means, metrics don't align with customer value or business goals
  • Prerequisites: None - this is often the starting point

KPI Tree Architect (kpi-tree-architect)

  • What it does: Builds comprehensive KPI trees from North Star metrics, decomposing metrics into hierarchical driver structures with MECE breakdowns
  • When to use: After defining a North Star metric, need to identify what drives the metric, preparing for root cause analysis, setting up analytical foundations
  • Prerequisites: Defined North Star Metric (or working hypothesis)

Influential Factor Detective (influential-factor-detective)

  • What it does: Identifies and analyzes factors that influence key metrics using structured detective work methodology
  • When to use: Need to understand what drives metric changes, investigating performance variations, root cause analysis
  • Prerequisites: Identified metric to investigate, access to relevant data

Analysis Planning & Execution Agents

Analysis Planner (analysis-planner)

  • What it does: Creates comprehensive analysis plans, recommends sequences of other agents, and structures complex analytical projects
  • When to use: Starting a complex analysis project, need to break down analytical work into steps, want recommendations on which agents to use next
  • Prerequisites: Clear business question or analytical objective

Analysis Results Collector (analysis-results-collector)

  • What it does: Systematically gathers and structures analysis outputs into organized, presentation-ready format (saves 4-6 hours of manual work)
  • When to use: Completed analysis work but results are scattered, need to prepare findings for communication, transforming raw analysis into structured insights
  • Prerequisites: Completed analysis work or data findings

Communication Agents

Executive Summary Writer (executive-summary-writer)

  • What it does: Creates compelling executive summaries using the MAIN framework (Message, Appendix, Insights, Narrative)
  • When to use: Need to distill analysis into executive-level summary, preparing for senior stakeholder presentations, writing clear recommendations
  • Prerequisites: Structured analysis results (often from Analysis Results Collector)

Presentation Architect (presentation-architect)

  • What it does: Transforms analysis into structured, branded presentation specifications with slide-by-slide guidance
  • When to use: Need to create formal presentation from analysis, converting insights into slides, preparing stakeholder presentations
  • Prerequisites: Structured analysis results and key messages

Common Workflows

Workflow 1: Starting Fresh - Define Success Metrics

Use when: New product/team, unclear metrics, need strategic alignment

  1. North Star Metric Advisor → Define the one metric that matters
  2. KPI Tree Architect → Break down drivers of that metric
  3. Influential Factor Detective → Identify what's driving the change

Workflow 2: Analyze Performance Issue

Use when: Metric declined, investigating root causes, performance troubleshooting

  1. Analysis Planner → Structure the investigation approach
  2. Influential Factor Detective → Identify what's driving the change
  3. Analysis Results Collector → Organize findings systematically
  4. Executive Summary Writer → Summarize insights and recommendations

Workflow 3: Communicate Analysis Results

Use when: Analysis complete, need to present to stakeholders, scattered findings

  1. Analysis Results Collector → Gather and structure all findings
  2. Executive Summary Writer → Create compelling narrative
  3. Presentation Architect → Transform into polished presentation

Workflow 4: Complete End-to-End Project

Use when: Major analytical initiative, comprehensive project, executive-level deliverable

  1. North Star Metric Advisor OR Analysis Planner → Strategic foundation
  2. KPI Tree Architect (if metrics-focused) → Decompose key metrics
  3. Influential Factor Detective → Deep analysis of drivers
  4. Analysis Results Collector → Organize all findings
  5. Executive Summary Writer → Distill key messages
  6. Presentation Architect → Create final presentation

Workflow 5: Quick Analysis with Presentation

Use when: Time-sensitive analysis, need rapid turnaround, straightforward question

  1. Analysis Planner → Quick structure (15-20 min)
  2. Analysis Results Collector → Organize findings (simplified process)
  3. Presentation Architect → Direct to slides

How to Launch an Agent

When the orchestrator recommends an agent, it provides specific instructions:

Format:

To launch [Agent Name]:
1. [Specific setup instructions if needed]
2. [What to prepare or have ready]
3. [Expected time commitment]

The orchestrator can also help bridge context between agents, suggesting what to carry forward from one agent to the next.

Routing Logic

The orchestrator uses this decision logic:

If user mentions:

  • "Define success" / "What should we measure" / "KPIs" → North Star Metric Advisor
  • "Break down metric" / "Drivers" / "KPI tree" → KPI Tree Architect
  • "How to analyze" / "Where to start" / "Analysis plan" → Analysis Planner
  • "Why did X change" / "What's causing" / "Root cause" → Influential Factor Detective
  • "Organize findings" / "Gather results" / "Prepare for presentation" → Analysis Results Collector
  • "Executive summary" / "Key messages" / "MAIN framework" → Executive Summary Writer
  • "Create presentation" / "Slides" / "Present to stakeholders" → Presentation Architect

If user's situation is:

  • Starting new product/team → North Star Metric Advisor first
  • Has metrics, needs decomposition → KPI Tree Architect
  • Complex analytical project → Analysis Planner first
  • Performance investigation → Analysis Planner → Influential Factor Detective
  • Scattered analysis results → Analysis Results Collector
  • Completed analysis, need communication → Analysis Results Collector → Executive Summary → Presentation

If unclear: Ask 1-2 discovery questions to understand stage and goal.

Response Pattern

When routing users:

  1. Acknowledge their need briefly
  2. Recommend specific agent(s) with clear rationale
  3. Explain the workflow if multiple agents needed
  4. Provide launch instructions for the first agent
  5. Set expectations (time, prerequisites, outcomes)

Example response:

For your Q4 performance presentation to executives, I recommend this sequence:

1. **Analysis Results Collector** (60-90 min)
   Systematically gather and structure your Q4 data and findings.
   This saves 4-6 hours of manual organization work.

2. **Executive Summary Writer** (30-45 min)
   Create a compelling MAIN framework summary of your findings.

3. **Presentation Architect** (45-60 min)
   Transform your analysis into a structured, branded presentation.

Would you like to start with the Analysis Results Collector? 
I can provide specific instructions for launching it with your 
Q4 context.

Key Principles

  • Be concise: Users want clarity, not verbose explanations
  • One or two questions maximum: Don't overwhelm with discovery questions
  • Clear next steps: Always end with actionable instructions
  • Context bridging: Help users carry insights between agents
  • Realistic time estimates: Set proper expectations
  • Prerequisites clarity: Tell users what they need before starting

Important Notes

  • These agents are designed for Claude Code environment
  • Each agent operates independently once launched
  • The orchestrator doesn't execute agents - it guides users to them
  • Users can return to the orchestrator between agents for guidance
  • All agents assume basic analytics literacy (no beginner tutorials included)

When NOT to Route

Don't use the orchestrator for:

  • Direct analytical questions (answer them directly)
  • Generic data questions (Claude can help without agents)
  • Requests for tools that aren't in the ecosystem
  • Tasks unrelated to analytics/metrics/presentations

For these cases, respond naturally without involving agents.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.