From aida-core
Create, optimize, validate, and list CLAUDE.md memory files across project, user, and plugin scopes (two-phase API), routed through /aida. Use for structured CLAUDE.md management. Do NOT use for a one-off manual edit to a single CLAUDE.md — edit that file directly.
How this skill is triggered — by the user, by Claude, or both
Slash command
/aida-core:claude-md-managerThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
<!-- SPDX-FileCopyrightText: 2026 The AIDA Core Authors -->
Manages CLAUDE.md configuration files that provide project-specific guidance to Claude Code. Supports the full CLAUDE.md hierarchy: project, user, and plugin scopes.
This skill activates when:
/aida claude create/aida claude optimize/aida claude validate/aida claude listGenerate a new CLAUDE.md file at the specified scope. Detects project context (languages, tools, commands) and renders from templates.
Audit an existing CLAUDE.md against best practices, generate findings with fix suggestions, and optionally apply fixes. Reports a quality score (0-100).
Run validation checks on existing CLAUDE.md files: structure, consistency, best practices, and alignment with detected project facts. Returns results without modifications.
Find all CLAUDE.md files in the hierarchy and report their validation status.
All operations follow the two-phase pattern:
Analyze context and return questions that need user input.
python {base_directory}/scripts/manage.py --get-questions \
--context='{"operation": "create", "scope": "project"}'
Returns:
{
"questions": [],
"inferred": {
"name": "my-project",
"languages": ["Python"],
"tools": ["Git", "pytest"],
"commands": [{"command": "make test", "description": "Run tests"}]
},
"existing": null
}
Execute the operation with provided context and responses.
python {base_directory}/scripts/manage.py --execute \
--context='{"operation": "create", "scope": "project"}' \
--responses='{"name": "my-project"}'
Returns:
{
"success": true,
"message": "Created CLAUDE.md at ./CLAUDE.md",
"path": "./CLAUDE.md"
}
| Scope | Path | Use Case |
|---|---|---|
project | ./CLAUDE.md | Project documentation |
user | ~/.claude/CLAUDE.md | Global user preferences |
plugin | .claude-plugin/CLAUDE.md | Plugin documentation |
Base Directory: Provided when skill loads via
<command-message> tags.
Script Execution:
{base_directory}/scripts/manage.py
| Script | Purpose |
|---|---|
scripts/manage.py | Two-phase API entry point |
scripts/operations/claude_md.py | Core CLAUDE.md operations |
scripts/operations/utils.py | Shared utility re-exports |
| Document | Purpose |
|---|---|
references/claude-md-workflow.md | End-to-end workflow guide |
references/best-practices.md | Best practices and scoring |
| Template | Purpose |
|---|---|
templates/claude-md/project.md.jinja2 | Project-scope template |
templates/claude-md/user.md.jinja2 | User-scope template |
templates/claude-md/plugin.md.jinja2 | Plugin-scope template |
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 aida-core/aida-core-plugin --plugin aida-core