From memplexor
Memplexor meta-skill — explains how project-local guidelines work. The real per-project guidelines live at .claude/skills/project-guidelines/SKILL.md and are generated by /setup or /memory-refresh. Use when a developer asks "how does the project-guidelines skill work" or wants to customize the template.
How this skill is triggered — by the user, by Claude, or both
Slash command
/memplexor:project-guidelinesThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
This is the **plugin-level** project-guidelines skill. It does not contain
This is the plugin-level project-guidelines skill. It does not contain
project-specific rules itself. Project-specific rules are generated by /setup
into the target project at:
.claude/skills/project-guidelines/SKILL.md
That generated file is what Claude actually consults when working in a project that has been initialized with Memplexor.
The generated SKILL.md must follow this shape:
---
name: project-guidelines
description: Project conventions for <project name> — <stack>. Apply before editing code in this repo.
---
Then markdown sections:
pnpm build, cargo build)pytest -q, npm test)src/, app/, tests/, etc. with one-line purposepackage.json, pyproject.toml, Cargo.toml, go.mod,
Gemfile, composer.json. Detect framework deps.package.json#scripts, Makefile, pyproject.toml
[tool.poetry.scripts], [tool.pdm.scripts], Cargo.toml [[bin]],
go.mod + common make targets..eslintrc*, .prettierrc*, tsconfig.json,
ruff.toml, mypy.ini, pyproject.toml[tool.ruff|mypy],
rustfmt.toml, .golangci.yml.node_modules/,
.git/, dist/, build/, target/).The plugin-level template lives at
${CLAUDE_PLUGIN_ROOT}/templates/project-guidelines.SKILL.md. Edit that file
and re-run /setup --refresh to regenerate.
/setup --refresh and /memory-refresh --skill only if it has the <!-- Generated by Memplexor --> marker on the
first line of the body.--force).npx claudepluginhub krzemienski/memplexor --plugin memplexorCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.