verification

Verifies end-to-end application flows across browser, API, and data layers with evidence.

Updated Aug 21, 2025
One-click install
npx skills add https://github.com/Adithiya-S/AI-Study-Companion --skill verification-adithiya-s
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verification
Source: https://github.com/Adithiya-S/AI-Study-Companion/tree/main/.agents/skills/verification
Command: npx skills add https://github.com/Adithiya-S/AI-Study-Companion --skill verification-adithiya-s

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When a feature "almost works" or a dev server just started, developers lack a systematic way to confirm the entire user story actually functions across every layer — UI, API routes, data sources, and response rendering — instead of checking each piece in isolation. ## Core Features & Use Cases - Story Inference: Reads recently edited files, routes, and environment files to infer the exact user story being built before running any checks. - Layer-by-Layer Evidence Gathering: Inspects browser console, dev server terminal, runtime logs, and environment variables, reporting findings at each boundary. - Data Flow Tracing: Walks the path from UI trigger through client request, API handler, external dependencies, and back to rendered response, catching common breaks like missing awaits, wrong HTTP methods, and absent env vars. - Use Case: After implementing a new form that submits to an API route backed by a database, ask for a full verification to confirm the browser renders, the request reaches the handler, the database call succeeds, and the response renders correctly. ## Quick Start Ask the assistant to verify the full flow end to end for the feature you just built.

Frequently Asked Questions about verification

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

FAQPage Schema
How do I verify my full-stack feature works end to end?

End-to-end verification traces the flow from the UI trigger through the client request, API route handler, and data source, then back to the rendered response. Each boundary is checked with concrete evidence such as console output, server logs, and response shapes.

What should I check when my app is not working correctly?

Start by inferring the user story, then gather evidence at each layer: browser console errors, dev server terminal output, runtime logs, and environment variables. Common breaks include missing awaits, wrong HTTP methods, and env vars absent at runtime.

Does end-to-end verification work with Next.js and Vite projects?

Yes, the verification flow triggers on dev servers started with next dev, vite, npm run dev, pnpm dev, and similar commands. It inspects route structures in app/ or pages/ directories and checks .env files for configuration issues.

Why does my API route work locally but fail after deployment?

This usually happens when environment variables exist in .env.local but are missing in the deployed runtime, or when server-only modules are imported in client components. Verification compares expected versus actual env vars and checks runtime logs for the exact failure.

When should I stop debugging a broken flow?

Stop at the first confirmed broken boundary, report it with evidence and a specific fix, and do not continue past the break. Also stop if two consecutive layers return no useful signal, which indicates an observability gap requiring added logging.