high-stakes-analytics-decision-lab
A platform-neutral analytical Skill that profiles messy data, selects case-adaptive methods, and produces source-backed visual reports for high-stakes decisions
What is high-stakes-analytics-decision-lab?
A platform-neutral analytical Skill that profiles messy data, selects case-adaptive methods, and produces source-backed visual reports for high-stakes decisions.
Overview of high-stakes-analytics-decision-lab
What it is
High-Stakes Analytics & Decision Lab is a platform-neutral Agent Skill and reproducible research portfolio for high-stakes decisions. It follows an evidence-gated sequence: evidence before method, readiness before analysis, adaptive routes, honest endpoints, dependent uncertainty, and traceable communication. The system is not a fixed report generator; it can stop, ask for a named decision, or produce a bounded analytical product without upgrading weak evidence into a stronger claim.
Key features
- Evidence-gated orchestration: stops, asks for a named decision, or produces a bounded analytical product - Data readiness gate: preserves the original source, establishes a contract, checks grain and keys, profiles quality and privacy, and produces a dry-run remediation plan with four statuses (ready, ready_with_documented_limitations, needs_user_confirmation, blocked)
- Four analytical routes: descriptive, diagnostic, predictive, prescriptive; no mandatory recommendation - Two products: Evidence Intelligence Report (primary record) and Decision Intelligence Brief (conditional, only when a separate decision layer is justified)
- Traceable communication: links claims and accessible figures to JSON, CSV, hashes, and rerunnable code - Platform-neutral: installable as an Agent Skill via npx - Public portfolio contains 15 primary reports, 10 conditional briefs (25 intelligence products), and 119 canonical accessible figures
Use cases
High-stakes analysis where questions are underspecified, data contracts are implicit, cleaning choices are hidden, uncertainty is treated as independent, or a recommendation is written because the template expects one. Example domains mentioned include operational demand, distribution shift, and scarce-capacity pilots.
Getting started
bash npx skills add limingrui679-design/high-stakes-analytics-decision-lab -g After installation, use the Skill by starting with a decision or evidence question. Direct repository entry points include python3 scripts/hsadl.py doctor for environment audit, python3 scripts/hsadl.py demo for a safe walkthrough, and other commands for routing, starting, validating, and running cases. See docs/getting-started.md for Codex-specific and no-install options.
License and pricing
MIT license, open source.
Based on the project's public README and repository metadata.
