Skip to content

florianbonnet14/ThePowerOfAnalytics_ClaudeSkills / analysis-results-collector

analysis-results-collector

Verified

Transform completed analysis work into structured, communication-ready documentation by conducting guided conversations that extract key findings, evidence, and recommendations. Use when a user has completed an analysis (typically following an analysis plan created by the analysis-planner skill) and needs to document results, create findings summaries, build executive reports, or prepare analysis outcomes for stakeholder communication. This skill systematically gathers what was tested, what was found, and what should be done next, then generates a professional markdown document ready for distribution.

references/conversation_guide.md

View on GitHub
# Conversation Patterns for Results Collection

This reference provides guidance on how to conduct effective conversations to gather analysis results.

## Core Principles

### 1. Progressive Disclosure
Don't overwhelm users with all questions at once. Gather information in logical phases:
- Phase 1: Understand which hypotheses were analyzed
- Phase 2: Determine which were validated/invalidated
- Phase 3: Collect detailed findings for each analyzed hypothesis
- Phase 4: Gather recommendations and next steps

### 2. Adaptive Questioning
Tailor follow-up questions based on:
- Whether hypothesis was validated or invalidated
- Complexity of the analysis performed
- Amount of detail already provided

### 3. Extract, Don't Transcribe
Transform raw analytical findings into communication-ready content:
- Remove jargon and technical details unless necessary
- Focus on "so what" - business implications
- Quantify impact wherever possible
- Translate data into decisions

## Phase 1: Hypothesis Triage

**Objective:** Quickly understand the scope of analysis completed

**Approach:**
- Present list of all hypotheses from the plan
- Ask user to identify which were analyzed vs not analyzed
- For "not analyzed" items, capture brief reason why

**Example questions:**
- "Looking at your analysis plan, I can see 8 hypotheses (H1-H8). Which of these did you actually analyze?"
- "For the hypotheses you didn't analyze (like H5, H6, H7), what was the reason? Was it that earlier hypotheses explained the issue, or something else?"

**What to capture:**
- Analyzed: [List of hypothesis IDs]
- Not analyzed: [List with brief reasons]

## Phase 2: Validation Status

**Objective:** Determine which analyzed hypotheses were proven true vs false

**Approach:**
- For each analyzed hypothesis, ask for validation status
- Use simple language: "Did this turn out to be true or false?"
- Accept nuanced answers ("partially validated")

**Example questions:**
- "For H1 (Account Setup Completion Rate Dropped), did your analysis validate this hypothesis or invalidate it?"
- "Was this hypothesis proven, disproven, or somewhere in between?"

**What to capture:**
- Status: Validated | Invalidated | Partially Validated
- Confidence level: High | Medium | Low
- Impact contribution: X% of total decline

## Phase 3: Detailed Findings Collection

**Objective:** Gather crisp, communication-ready summaries of what was found

**Critical distinction:** We're not asking users to paste raw data or reproduce their analysis. We're asking them to summarize findings in a way that can be communicated to stakeholders.

### For VALIDATED Hypotheses

**What we need:**
1. **Core finding** (2-3 sentences): What actually happened?
2. **Supporting evidence** (3-5 key data points): What proves this?
3. **Impact quantification**: How much did this contribute?
4. **Business implication**: Why does this matter?

**Example questions:**
- "You validated that account setup completion dropped. Can you summarize what you found in 2-3 sentences that I could share with leadership?"
- "What are the 3-4 most important data points that support this finding?"
- "Can you quantify the impact? For example, did setup completion drop from X% to Y%, and how much of the overall 15% decline does this explain?"
- "Why does this matter to the business? What's the implication?"

**Adaptive follow-ups based on detail level:**
- If user provides raw numbers only → Ask for interpretation
- If user provides technical jargon → Ask for plain language version
- If user provides long explanation → Ask for executive summary version
- If user provides conclusion without evidence → Ask for key supporting data points

### For INVALIDATED Hypotheses

**What we need:**
1. **What was ruled out** (1-2 sentences): What did we learn wasn't the cause?
2. **Key evidence** (1-2 data points): What disproved this?
3. **Value of ruling this out**: Why was this useful to test?

**Example questions:**
- "You invalidated the user mix hypothesis. In plain language, what did you learn?"
- "What specific evidence showed this wasn't the cause?"
- "Was this hypothesis worth testing, or should it have been skipped?"

### For PARTIALLY VALIDATED Hypotheses

**What we need:**
1. **What was true vs false**: Which parts validated?
2. **Partial impact**: How much did the validated portion contribute?
3. **Remaining questions**: What's still unclear?

**Example questions:**
- "You said this was partially validated. Which aspects were true and which weren't?"
- "For the parts that were validated, what's the impact?"
- "What questions remain unanswered?"

## Phase 4: Recommendations

**Objective:** Capture actionable next steps with clear prioritization

**What we need:**
1. **Recommendation title**: Clear, action-oriented name
2. **Root cause addressed**: Which finding does this fix?
3. **Expected impact**: Projected improvement
4. **Effort required**: Resource/time estimate
5. **Specific actions**: 3-5 concrete steps
6. **Success metrics**: How to measure if it worked

**Example questions:**
- "Based on these findings, what are your top 3-5 recommendations?"
- "For each recommendation, what's the expected impact on the North Star Metric?"
- "What are the specific action items for [Recommendation 1]?"
- "How will you measure success for this recommendation?"

**Prioritization questions:**
- "If you could only implement one recommendation, which would it be and why?"
- "Which recommendations are quick wins vs longer-term investments?"

## Handling Common Scenarios

### Scenario: User Provides Too Much Detail
**Response:** "That's helpful technical detail. For the final document, can you distill this into 2-3 key points that would resonate with leadership?"

### Scenario: User Provides Too Little Detail
**Response:** "Can you elaborate a bit more? For example, what specific data points support this finding?"

### Scenario: User Is Uncertain About Impact
**Response:** "Even a rough estimate is helpful. Would you say this contributed roughly X% to the decline, or is it more/less than that?"

### Scenario: User Jumps Around Topics
**Response:** "Great information! Let me make sure I capture this properly. Let's finish discussing [current hypothesis] first, then we can move to [other topic]."

### Scenario: User Wants to Paste Entire Analysis
**Response:** "I appreciate you have detailed analysis. Rather than paste everything, let's work through the key findings together so I can capture them in a format that's ready for communication. What was the single most important thing you learned from this hypothesis?"

## Output Quality Standards

Before moving to next hypothesis, ensure you have captured:
- ✅ Clear, jargon-free summary of finding
- ✅ 3-5 specific data points or evidence
- ✅ Quantified impact where possible
- ✅ Business implication stated
- ✅ Information is communication-ready (not raw data dump)

## Progressive Writing Approach

After gathering information for 2-3 hypotheses:
- Pause to write up those sections
- Show user draft sections for feedback
- Refine based on their input
- Then continue with remaining hypotheses

This allows:
- Early validation of format and style
- Course correction if needed
- Avoids losing information in long conversations
- Keeps user engaged with tangible progress
rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file).Stars: 198kUpdated Claude CodeCodexUnverifiedInstall
doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.Stars: 198kUpdated Claude CodeCodexVerifiedInstall