From superskills
Discover an existing project's conventions and generate or refresh minimal AI spec files (.superskills/conventions.md, AGENTS.md, CLAUDE.md). Use when a project lacks AGENTS.md/CLAUDE.md, when the user asks to generate conventions ("生成规范", "init ai docs", "discover conventions"), or when conventions are reported stale.
How this skill is triggered — by the user, by Claude, or both
Slash command
/superskills:discoverThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Generate or refresh the project's AI-facing spec files. Keep everything short — these files load into every session, so every line costs tokens forever.
Generate or refresh the project's AI-facing spec files. Keep everything short — these files load into every session, so every line costs tokens forever.
git log --oneline -20 for the commit message convention.superskills/conventions.md — single source of truth, max 80 lines:
Every line must be project-specific and evidenced. Drop generic advice ("write clean code"). Prefer one line over three.
AGENTS.md (if missing) — max 20 lines: one-paragraph project description, key commands, then exactly these pointers:
Read .superskills/conventions.md before writing code.Check .superskills/learnings/INDEX.md for past learnings; read a linked entry when relevant.If AGENTS.md exists, only append the two pointer lines when missing. Never rewrite existing user content.
CLAUDE.md (if missing) — exactly these imports:
@AGENTS.md
@.superskills/conventions.md
If it exists, append missing imports only.
.superskills/learnings/INDEX.md — create with a # Learnings header if missing.
If .superskills/conventions.md already exists:
.superskills/learnings/ that have hardened into stable conventions directly into conventions.md, then delete those learning files and their INDEX lines.Finish by suggesting a commit of .superskills/, AGENTS.md, and CLAUDE.md.
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub mrlyk/superskills --plugin superskills