drift-check

Compare an investigation's current state against an earlier snapshot and report changes.

14|Updated Jun 29, 2026
One-click install
npx skills add https://github.com/sparkline-ventures/product-eval --skill drift-check-sparkline-ventures
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: drift-check
Source: https://github.com/sparkline-ventures/product-eval/tree/main/skills/drift-check
Command: npx skills add https://github.com/sparkline-ventures/product-eval --skill drift-check-sparkline-ventures

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you compare an investigation’s current state against an earlier snapshot so you can see whether the problem framing, evidence base, and decision confidence have genuinely changed over time.

Core Features & Use Cases

  • Snapshot comparison: Compares the present investigation against a prior snapshot or git ref to identify what moved.
  • Evidence and score tracking: Highlights evidence added or retired, along with Value and Confidence score changes and their drivers.
  • Narrative and consistency review: Detects drift in the problem statement, persona, scope, and decision narrative, and flags consistency issues like vanity-metric creep or scope mismatch.
  • Use case: A product team wants to know whether last month’s research still supports the same bet, or whether the decision memo needs to be revised before re-approval.

Quick Start

Ask the skill to compare the current investigation against a prior snapshot and summarize what changed, including evidence, scores, narrative drift, and confidence trajectory.

Frequently Asked Questions about drift-check

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I compare an investigation's current state against an earlier snapshot?

To compare an investigation's current state against an earlier snapshot, you perform snapshot diffing to identify changes in problem framing, evidence, and confidence scores over time. This analysis reports what moved between the two versions.

What is decision drift and how do I track it in a research project?

Decision drift occurs when a project's problem statement, scope, or narrative framing changes over time. You can track decision drift by applying cross-stage consistency checks against a prior snapshot to detect vanity-metric creep and scope mismatches.

How do I check if evidence and confidence scores have changed in a git history?

To check if evidence and confidence scores have changed in a git history, you compare your current investigation against a prior git ref. This highlights evidence added or retired, Value and Confidence score changes, and their underlying drivers.

When do I need to run a confidence trajectory analysis for a decision log?

You need to run a confidence trajectory analysis for a decision log when verifying if prior research still supports the same bet. It evaluates confidence trajectory changes across problem statements, evidence, themes, and scores to determine if a decision memo needs revision.

Can I detect narrative drift and consistency issues using a snapshot diff?

Yes, you can detect narrative drift and consistency issues using a snapshot diff. This approach flags drift in the problem statement, persona, and scope, while identifying consistency issues like vanity-metric creep or scope mismatch across your investigation.