skills.rest
greenflash-ai
greenflash-aiOFFICIAL
Explore all skills by this author →
WebsiteGitHub

greenflash-diagnose

Diagnose root causes of silent evaluation failures and surface failing tools with fixes.

Updated Mar 25, 2026
Search Tags:#debugging#diagnose#guardrail#root-cause#code-change#prompt-fix#eval-failures
One-click install
npx skills add https://github.com/greenflash-ai/agent-skills --skill greenflash-diagnose
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: greenflash-diagnose
Source: https://github.com/greenflash-ai/agent-skills/tree/main/skills/greenflash-diagnose
Command: npx skills add https://github.com/greenflash-ai/agent-skills --skill greenflash-diagnose

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Root-cause analysis for silent eval failures, surfacing failing tools, guardrail violations, and user friction with concrete fixes that can be implemented directly in your codebase.

Core Features & Use Cases

  • Surface failing tools with confidence scores and affected user segments
  • Identify root causes, guardrail violations, and friction patterns across products and conversations
  • Provide actionable fixes and prompts changes ready for one-click deployment
  • Use Case: Imagine a product with several evals returning vague errors; this skill diagnoses the root cause and suggests precise code or prompt changes

Quick Start

Run diagnostics on a project to identify failing tools and friction, then apply the recommended fixes.

Frequently Asked Questions about greenflash-diagnose

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

FAQPage Schema
How do I diagnose silent evaluation failures in my codebase?▼

Diagnose silent evaluation failures by running diagnostics on your local project to surface failing tools, guardrail violations, and friction patterns with confidence scores and recommended fixes.

What is the best way to find root causes for inconsistent prompts and user drop-offs?▼

Root-cause analysis for inconsistent prompts and user drop-offs identifies affected user segments, guardrail violations, and friction patterns, providing concrete code or prompt changes ready for deployment.

What do I need to run diagnostics and apply fixes to failing tools?▼

Running diagnostics requires access to your local project, the Greenflash diagnostics API, and authorization to modify prompts, code, or tool configurations to apply the recommended remediation.

Why do my evaluations return vague errors and how can I fix them?▼

Evaluations returning vague errors are diagnosed by surfacing failing tools and root causes with confidence scores, generating precise code or prompt changes for one-click deployment.

Can I use this to fix guardrail violations across multiple products?▼

Yes, diagnosing multi-product environments identifies failing tools, friction, and guardrail violations across products and conversations, providing actionable remediation with evidence and affected user data.

Related Skills

Explore More Software Engineering Skills→
greenflash-aigreenflash-ai

greenflash-verify

Verify Greenflash integration configuration and data flow for Python and TypeScript projects.

Official
Advanced
JaimeJunrJaimeJunr
2

run-tests

Execute unit and integration tests, analyze failures, and re-run to verify corrections.

Community
Advanced
lebsrallebsral
11

ai-monitoring

Monitor AI model performance, drift, and cost in production.

Community
Advanced
knckknckknckknck

wcag-audit-patterns

Audit WCAG 2.2 conformance with axe-core, pa11y, and lighthouse.

Community
Advanced