workflow-data-outlier-detection
WHAT: Statistical outlier detection across financial datasets — univariate (z-score, IQR, p1/p99), cross-sectional (peer-relative), time-series (moving-average spike), and factor-residual methods. Outliers are flagged with a review_flag column; they are never silently deleted. WHEN: Invoke when a raw numeric dataset has been ingested and must be screened for implausible values before entering a financial model, or when a model run has produced implausible output and the upstream data feed is suspect.
Source facts
- Repository
- fall-development-rob/corp_finance
- Last source activity
- May 11, 2026 at 14:10
- Detected SKILL.md language
- English
- Stars
- 8
- Forks
- 2
Install options
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Review the source files
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Showing SKILL.md
| name | workflow-data-outlier-detection |
|---|---|
| description | WHAT: Statistical outlier detection across financial datasets — univariate (z-score, IQR, p1/p99), cross-sectional (peer-relative), time-series (moving-average spike), and factor-residual methods. Outliers are flagged with a review_flag column; they are never silently deleted. WHEN: Invoke when a raw numeric dataset has been ingested and must be screened for implausible values before entering a financial model, or when a model run has produced implausible output and the upstream data feed is suspect. |