Urgent.News

What's breaking now, across thousands of outlets.

AI

Every Claude Code skill I maintain, in one repository

Most of what I ask Claude Code to do at work is the same few jobs: check an AWS account, review a Conditional Access setup, threat model an agent, build an evidence pack for an auditor, tidy a repository. Over the year each of those became a skill, and the skills grew into eleven repositories. This article is about the one repository that now holds all of them, and about where it stops. Why one…

Claude Code, a tool used for various tasks at work, has traditionally maintained a separate repository for each skill it performs. Over time, as more skills were added, this resulted in 11 individual repositories. The author then decided to consolidate all these skills into a single repository called "claude-skills." This repository now contains 125 skills spread across 16 plugins, seamlessly copied from the original 11 sources and kept synchronized with them.

The consolidated repository, claude-skills, includes skills for agent security, agent identity governance, AWS security, Microsoft 365 governance, compliance evidence, everyday development, repository engineering, ways of working, Claude Code setup, GitHub management, and Mac maintenance. Each skill is accompanied by a detailed description, usage instructions, and limitations.

The repository also ensures honesty by regularly syncing changes from the source repositories and maintaining a detailed log of which source repository each skill originated from.

To install the repository, users can run a single command in Claude Code: /plugin marketplace add basitalisandhu/claude-skills /plugin install aws-security@claude-skills. Alternatively, users can clone the repository from GitHub and run a Python installer script. The installer is lightweight Python code without any dependencies. Other agents that support the Agent Skills format can also install the skills using npx skills add basitalisandhu/claude-skills.

While the repository provides a comprehensive set of skills, it is important to note that these skills operate offline and do not make any network calls or send telemetry. Additionally, the bundled scripts are standard-library Python, and every skill description includes a "Limits" section to clarify what the skill cannot do. If there are any issues or bugs, users are encouraged to open an issue or a pull request on the respective source repository, or report the problem on the claude-skills repository itself.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at dev.to →

More in AI

VAI Assistant: a non-injection AI customer service tool for WeChat on Windows

VAI助手 (formerly "微信管家", https://vx.cuxn.cn/ ) is a Windows desktop app that auto-answers routine customer questions in WeChat for online sellers, WeChat merchants and private-domain customer service…

  • VAI Assistant is a Windows desktop tool for WeChat customer service.
  • Uses local RAG knowledge base from PDF/DOCX files.
  • Reads messages, types replies at system UI level, not injecting into WeChat.

Why Your AI Agent Produces Duplicates (It’s a Modeling Error, Not a Logic Error)

Most duplicate-dispatch bugs in AI agents aren’t caused by missing retry guards or broken idempotency keys. They’re caused by collapsing two independent questions into one state field.

  • Duplicate-dispatch bugs stem from modeling error, not logic error.
  • Two-Axis Model separates DeliveryState and AcceptanceState to prevent duplicates.
  • ShouldRedispatch() function determines dispatch retries based on state conditions.

More from Tuesday 13 October →