Data Analysis Litiao

Turn raw data into decisions with statistical rigor, proper methodology, and awareness of analytical pitfalls.

Security Scan

This is an instruction-only data‑analysis guidance skill whose files and runtime instructions match the stated purpose and do not request credentials, binaries, or perform I/O — the main issue is minor metadata inconsistencies (missing/unknown source and mismatched metadata) that warrant provenance checks before trusting it in production.

Purpose & Capability

The skill's name, description, and the SKILL.md / techniques/pitfalls content are coherent: they provide analysis methodology, checks, and recommended outputs appropriate for a 'data analysis' helper. However, metadata inconsistencies exist: the registry metadata ownerId (kn7838z...) and slug (data-analysis-litiao) do not match the _meta.json ownerId (kn73vp5...) and slug (data-analysis). There is also no homepage or source URL listed. These mismatches are a provenance concern but do not indicate malicious functionality.

Instruction Scope

SKILL.md and the supporting docs are purely prescriptive about analytical methodology and outputs. They do not instruct the agent to read arbitrary system files, call external endpoints, exfiltrate data, or access environment variables. There is no step that broadens scope beyond data-analysis guidance.

Install Mechanism

No install spec and no code files beyond markdown—this is instruction-only. Nothing is downloaded or written to disk by an installer, which minimizes installation risk.

Credentials

The skill declares no required environment variables, credentials, or config paths. The instructions do not reference secrets or unrelated service tokens. The requested access is proportionate to its stated purpose.

Persistence & Privilege

Skill flags are default (not always:true), and autonomous invocation is enabled (the platform default). It does not request permanent presence nor system configuration changes. No concerning privilege escalation is requested.

Assessment

This skill appears to be a benign, instruction-only data-analysis helper. Before installing or allowing it to run on sensitive data: 1) verify provenance — ask the publisher for a homepage, source repository, or author identity because _meta.json and registry metadata disagree; 2) avoid feeding sensitive PII or credentials into examples run with the skill; 3) test on small, non-sensitive datasets to confirm outputs meet your expectations; and 4) treat its recommendations as guidance (not authoritative code) — validate analyses and check assumptions before making decisions based on its output.

When to Load

User asks about: analyzing data, finding patterns, understanding metrics, testing hypotheses, cohort analysis, A/B testing, churn analysis, statistical significance.

Core Principle

Analysis without a decision is just arithmetic. Always clarify: What would change if this analysis shows X vs Y?

Methodology First

Before touching data:

1. What decision is this analysis supporting?
2. What would change your mind? (the real question)
3. What data do you actually have vs what you wish you had?
4. What timeframe is relevant?

Statistical Rigor Checklist

• Sample size sufficient? (small N = wide confidence intervals)
• Comparison groups fair? (same time period, similar conditions)
• Multiple comparisons? (20 tests = 1 "significant" by chance)
• Effect size meaningful? (statistically significant ≠ practically important)
• Uncertainty quantified? ("12-18% lift" not just "15% lift")

Analytical Pitfalls to Catch

Pitfall	What it looks like	How to avoid
Simpson's Paradox	Trend reverses when you segment	Always check by key dimensions
Survivorship bias	Only analyzing current users	Include churned/failed in dataset
Comparing unequal periods	Feb (28d) vs March (31d)	Normalize to per-day or same-length windows
p-hacking	Testing until something is "significant"	Pre-register hypotheses or adjust for multiple comparisons
Correlation in time series	Both went up = "related"	Check if controlling for time removes relationship
Aggregating percentages	Averaging percentages directly	Re-calculate from underlying totals

For detailed examples of each pitfall, see pitfalls.md.

Approach Selection

Question type	Approach	Key output
"Is X different from Y?"	Hypothesis test	p-value + effect size + CI
"What predicts Z?"	Regression/correlation	Coefficients + R² + residual check
"How do users behave over time?"	Cohort analysis	Retention curves by cohort
"Are these groups different?"	Segmentation	Profiles + statistical comparison
"What's unusual?"	Anomaly detection	Flagged points + context

For technique details and when to use each, see techniques.md.

Output Standards

1. Lead with the insight, not the methodology
2. Quantify uncertainty — ranges, not point estimates
3. State limitations — what this analysis can't tell you
4. Recommend next steps — what would strengthen the conclusion

Red Flags to Escalate

• User wants to "prove" a predetermined conclusion
• Sample size too small for reliable inference
• Data quality issues that invalidate analysis
• Confounders that can't be controlled for