Skip to content
Agent Memory
Skill

/codebase-analyzer

This skill should be used when the user asks to "initialize auto-memory", "create CLAUDE.md", "set up project memory", or runs the /auto-memory:init command. Analyzes codebase structure and generates CLAUDE.md files using the exact template format with AUTO-MANAGED markers.

BOOST
From plugin
auto-memory
1582 skills1 agent4 commands4 hooks
Install
$ npx -y skills add severity1/claude-code-auto-memory --skill codebase-analyzer --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.
  • Slash command/codebase-analyzer

Context preview

The summary Claude sees to decide when to auto-load this skill.

This skill should be used when the user asks to "initialize auto-memory", "create CLAUDE.md", "set up project memory", or runs the /auto-memory:init command. Analyzes codebase structure and generates CLAUDE.md files using the exact template format with AUTO-MANAGED markers.

SKILL.md

codebase-analyzer.SKILL.md
name: codebase-analyzer
description: This skill should be used when the user asks to "initialize auto-memory", "create CLAUDE.md", "set up project memory", or runs the /auto-memory:init command. Analyzes codebase structure and generates CLAUDE.md files using the exact template format with AUTO-MANAGED markers.

Codebase Analyzer

Analyze project structure and generate CLAUDE.md files interactively.

Guidelines

**MANDATORY**: All rules below must be followed exactly. Violations produce incorrect CLAUDE.md content.

@../shared/references/guidelines.md

Algorithm

1. Check Existing CLAUDE.md

If CLAUDE.md already exists, ask the user how to handle it:

  • **Migrate**: Convert to auto-managed format (add markers)
  • **Backup**: Create CLAUDE.md.backup and generate fresh
  • **Merge**: Keep manual sections, add auto-managed sections
  • **Cancel**: Abort initialization

2. Scan Directory Structure

Detect frameworks and build systems:

  • `package.json` - Node.js/JavaScript
  • `pyproject.toml`, `setup.py` - Python
  • `Cargo.toml` - Rust
  • `go.mod` - Go
  • `Makefile` - Make-based builds
  • `Dockerfile` - Container builds

Extract build/test/lint commands from config files.

3. Identify Subtree Candidates

Look for framework boundaries that warrant separate CLAUDE.md files:

  • `src/` with 10+ source files
  • `lib/` directory
  • `packages/*` (monorepo packages)
  • `apps/*` (monorepo applications)

4. Detect Code Patterns

Analyze source files for conventions:

  • **Naming**: PascalCase, camelCase, snake_case usage
  • **Imports**: ES6 modules, CommonJS, ordering patterns
  • **Architecture**: Feature-based, layered, MVC patterns
  • **Style**: Indentation, quotes, semicolons

5. Fetch Memory Guidelines (Optional)

If network available, fetch from `https://code.claude.com/docs/en/memory`:

  • Use WebFetch tool
  • Extract relevant sections for context
  • Reference official guidelines above as primary source

6. Present Findings

Use AskUserQuestion to confirm:

  • Detected framework and commands
  • Suggested subtree locations
  • Detected patterns

7. Generate Memory Files

Read `.claude/auto-memory/config.json` to determine which files to generate based on `memoryFiles`:

| `memoryFiles` value | Files to generate | |---|---| | absent or `["CLAUDE.md"]` | `CLAUDE.md` only (full content) | | `["AGENTS.md"]` | `AGENTS.md` only (full content) | | `["CLAUDE.md", "AGENTS.md"]` | `AGENTS.md` (full content) + `CLAUDE.md` (redirect) |

Generate full-content files using the EXACT template structure. Follow these steps precisely:

1. **Copy template skeleton** - Use the appropriate template (CLAUDE or AGENTS) as the base structure 2. **Use exact marker format** - See Marker Syntax section below 3. **Replace placeholders** - Substitute `{{PLACEHOLDER}}` with detected content 4. **Include all required sections** - Even if content is minimal, include the section 5. **Add MANUAL section** at the end for user notes 6. **Size limits**: Root 150-200 lines, Subtrees 50-75 lines

When both `CLAUDE.md` and `AGENTS.md` are configured, also generate redirect files:

  • At the root: write `CLAUDE.md` using `CLAUDE.redirect.md.template` (static pointer, no markers)
  • For each subtree that gets an `AGENTS.md`: write a matching `CLAUDE.md` redirect alongside it

Marker Syntax

**CRITICAL**: Use the EXACT marker format below. Do NOT use variations.

<!-- AUTO-MANAGED: section-name -->
## Section Heading

Content goes here

<!-- END AUTO-MANAGED -->

For user-editable content:

<!-- MANUAL -->
## Custom Notes

Add project-specific notes here. This section is never auto-modified.

<!-- END MANUAL -->

**Common mistakes to avoid**:

  • `<!-- BEGIN AUTO-MANAGED: name -->` - WRONG (no BEGIN prefix)
  • `<!-- END AUTO-MANAGED: name -->` - WRONG (no name in closing tag)
  • `<!-- AUTO-MANAGED section-name -->` - WRONG (missing colon)

Section Definitions

Root CLAUDE.md Sections

Generate these sections in order:

| Section Name | Heading | Required | Placeholder | Content | |--------------|---------|----------|-------------|---------| | `project-description` | ## Overview | Yes | `{{DESCRIPTION}}` | Project name, tagline, key features | | `build-commands` | ## Build & Development Commands | Yes | `{{BUILD_COMMANDS}}` | Build, test, lint, run commands | | `architecture` | ## Architecture | Yes | `{{ARCHITECTURE}}` | Directory tree, key files, data flow | | `conventions` | ## Code Conventions | Yes | `{{CONVENTIONS}}` | Naming, imports, formatting rules | | `patterns` | ## Detected Patterns | Yes | `{{PATTERNS}}` | AI-detected recurring code patterns | | `git-insights` | ## Git Insights | No | `{{GIT_INSIGHTS}}` | Key commits, design decisions | | `best-practices` | ## Best Practices | No | `{{BEST_PRACTICES}}` | From official Claude Code docs |

Subtree CLAUDE.md Sections

Generate these sections in order:

| Section Name | Heading | Required | Placeholder | Content | |--------------|---------|----------|-------------|---------| | `module-description` | ## Purpose | Yes | `{{DESCRIPTION}}` | Module purpose and responsibility | | `architecture` | ## Module Architecture | Yes | `{{ARCHITECTURE}}` | Module structure, key files | | `conventions` | ## Module-Specific Conventions | Yes | `{{CONVENTIONS}}` | Module-specific rules | | `dependencies` | ## Key Dependencies | Yes | `{{DEPENDENCIES}}` | Module dependencies |

Templates

Reference the template files for exact structure:

CLAUDE.md Root Template

@templates/CLAUDE.root.md.template

CLAUDE.md Subtree Template

@templates/CLAUDE.subtree.md.template

AGENTS.md Root Template

@templates/AGENTS.root.md.template

AGENTS.md Subtree Template

@templates/AGENTS.subtree.md.template

CLAUDE.md Redirect Template (when both files configured)

@templates/CLAUDE.redirect.md.template

User Interactions

Use AskUserQuestion for: 1. Existing CLAUDE.md handling (migrate/backup/merge/cancel) 2. Subtree location confirmation 3. F

Read more
Ships withauto-memory

Your CLAUDE.md, always in sync. Minimal tokens. Zero config. Just works. A Claude Code plugin that watches what Claude Code edits, deletes, and moves - then quietly updates your project memory in the background. No manual maintenance needed.

Get the whole plugin
Stats
158
Stars
16
Forks
Quiet
Maintenance
Python
Language
MIT
License
6mo ago
Last commit
11mo ago
Created
43m ago
Added

Repo: severity1/claude-code-auto-memory

Other skills on auto-memory.