lading-optimize-rescue

Generate missing benchmarks and validate optimization claims in a structured rescue workflow.

98|16|Updated Mar 3, 2021
One-click install
npx skills add https://github.com/DataDog/lading --skill lading-optimize-rescue
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: lading-optimize-rescue
Source: https://github.com/DataDog/lading/tree/main/.claude/skills/lading-optimize-rescue
Command: npx skills add https://github.com/DataDog/lading --skill lading-optimize-rescue

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Optimization work often proceeds without solid benchmarks, making claims difficult to verify and slowing progress. This Skill provides a structured rescue workflow to generate missing evidence and validate optimization results.

Core Features & Use Cases

  • Phase-driven rescue process: pre-flight, audit, triage, evidence generation, validation, and recording.
  • Standardized outcomes: kept, discarded, and bugs_found with a controlled review cycle.
  • Use Case: when a proposed optimization lacks benchmarks, run the rescue to confirm improvements or identify regressions.

Quick Start

Use the lading-optimize-rescue to begin a rescue session, then follow the Phase steps documented in SKILL.md to generate evidence and validate results.

Frequently Asked Questions about lading-optimize-rescue

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

FAQPage Schema
How do I validate optimization claims that lack benchmarks?

A rescue workflow salvages optimization work by guiding you through audit, triage, evidence generation, and validation phases to generate missing benchmarks. This deterministic process categorizes outcomes as kept, discarded, or bugs found, ensuring claims are reproducibly verified.

What is a rescue workflow for optimization proposals?

A rescue workflow is a phase-driven process that salvages unverifiable optimization work by generating missing evidence through deterministic experiments. It progresses through pre-flight, audit, triage, validation, and recording phases to establish reproducible benchmark outcomes.

How do I ensure reproducibility when testing optimization results?

Reproducibility in optimization testing is ensured by enforcing deterministic experiments during evidence generation. The rescue workflow establishes strict outcome categories and integrates with validation cycles to confirm improvements or identify regressions consistently across runs.

When should I use a structured rescue process for bug discovery?

Use a structured rescue process when proposed optimizations lack solid benchmarks and claims become difficult to verify. This workflow applies deterministic experiments to audit proposals, generating evidence that either confirms improvements or surfaces hidden regressions and bugs.

What are the standard outcome categories for optimization validation?

The standard outcome categories for optimization validation are kept, discarded, and bugs found. These categories are established during the recording phase to classify whether a proposed optimization improves performance, fails to deliver, or introduces regressions requiring further investigation.