Skip to content
Back to skills

Connect

ASecurity

Use when the person is not connected, asks how connection works, or hits a connection failure.

  • 8 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added September 19, 2026
developmentapi

Works with

  • api

Security analysis

A100/100

Scanned September 19, 2026

npx -y skills add shoto290/kiroshi-monorepo --skill connect --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Connect?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Connect
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/shoto290-connect/badge)](https://www.skillsdirectory.com/skills/shoto290-connect)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: "connect"
description: "Use when the person is not connected, asks how connection works, or hits a connection failure."
---

Three sources, in this order: an existing Claude account already signed in on the machine (nothing to do, name the account and plan when the app exposes them), Sign in with Claude, an API key or a subscription token as a second field on the same screen.

Sign in with Claude is three steps: the app opens the browser, the person finishes there, the person pastes back the code the browser ends on. That paste is the step every sign-in ends on, and the screen carries the same link and the field for the code the whole time.

The five failures and the one gesture each: not signed in (sign in), key or token invalid, 401 (paste it again or sign in), session expired (sign in again), credit or usage limit (wait or top up on [claude.ai](https://claude.ai), nothing to fix here), plan that does not cover the usage (arrives at the first answer, never at login: change the plan or use a key).

One failure, one gesture, never a list of causes.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.