I Collected 208 Skills for Coding Agents. Here’s How I Stop Them Fighting Each Other

Description and use

The author assembled 208 skills for coding agents and resolved conflicts by applying clear trigger rules and limiting active skills to three per task.

Problem

With over 200 skills installed, multiple skills can fire simultaneously, causing conflicts and inefficiency.

Outcome

The curated skill set enables coding agents to execute tasks efficiently by preventing skill conflicts and ensuring only relevant skills are loaded.

Workflow

• Collect and edit 208 skill folders, each containing a SKILL.md file with a short description and instructions.
• Rewrite skill descriptions to act as precise triggers rather than marketing copy.
• Apply rules: load a skill only on clear description match, limit to 3 skills per task, prefer more specific matches, and avoid forced skills.
• Clone the repository: git clone https://github.com/Rachit-Sharma08/coding-agent-skills.git. Copy desired skill folders into ~/.claude/skills/ for Claude Code or ~/.agents/skills/ for OpenCode. Follow setup steps in SETUP.md.
• Use only the five skills needed for a task to keep the agent’s context small.

Evidence

I collected, edited and wired them together. Many are adapted from public work, including Anthropic's official skills, obra/superpowers, Corey Haines' marketing skills and OWASP cheat sheets.

Source

With 200+ skills installed, the problem is not "no skill". It is three skills firing at once.

Source

My rules: Load a skill only when the request clearly matches its description, not on one shared keyword. Max 3 skills per task. If two match, take the more specific one. No match? Work without a skill.

Source