From claude-code-project-bootstrap
Use when user wants to check an existing project against Claude Code workflow best practices, find missing hooks, weak guardrails, or gaps in CLAUDE.md configuration. Triggered by /audit-project command.
How this skill is triggered — by the user, by Claude, or both
Slash command
/claude-code-project-bootstrap:audit-projectThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
You are auditing the current project's Claude Code setup against best practices from the `claude-code-project-bootstrap` skill. Check every item, report findings, and offer to fix gaps.
You are auditing the current project's Claude Code setup against best practices from the claude-code-project-bootstrap skill. Check every item, report findings, and offer to fix gaps.
Run through ALL checks below. For each, report one of:
.git/ directory exists (project is a git repo)git remote -v)main branch directly (should be on a feature branch).gitignore exists.gitignore includes .claude/settings.local.json.gitignore includes .env and secret patterns (*.key, *.pem).gitignore includes OS files (.DS_Store, Thumbs.db).gitignore includes stack-specific entries (node_modules, pycache, target/, etc.)README.md existsCheck .claude/hooks/ directory:
validate-bash.sh exists and is executable
protect-files.sh exists and is executable
.env* filescredentials.json, secrets.json, *.key, *.pem)build-check.sh exists and is executable
package.json)Cargo.toml)go.mod)pyproject.toml/setup.py)*.xcodeproj/*.xcworkspace)scan-secrets.sh exists and is executable
sk-, AKIA, ghp_, gho_, glpat-)API_KEY=, PASSWORD=, etc.)-----BEGIN.*PRIVATE KEY)session-check.sh exists and is executable
.claude/hooks/ directory existsCLAUDE.md existsauto-format.sh exists and is executable
.ts/.tsx/.js/.jsx (prettier).py (ruff/black).rs (rustfmt).swift (swiftformat).go (gofmt)command -v guards (doesn't fail if formatter not installed)Check .claude/settings.json:
PreToolUse hook for Bash → validate-bash.shPreToolUse hook for Write|Edit → protect-files.shPostToolUse hook for Write|Edit → scan-secrets.shPostToolUse hook for Write|Edit → auto-format.shSessionStart hook → session-check.sh$CLAUDE_PROJECT_DIR (portable, not hardcoded)git ls-files .claude/settings.local.json should return empty)Check CLAUDE.md at project root:
Quick check for common security issues:
.env files committed (git ls-files '*.env*' should return empty)git ls-files '*credentials*' '*secrets*' '*.key' '*.pem')After running all 45+ checks, present results as a table:
## Audit Results
| Area | Status | Details |
|-----------------|--------|----------------------------------|
| Git & GitHub | PASS | Repo configured, .gitignore good |
| README | WARN | Missing contributing section |
| Hooks | FAIL | Missing scan-secrets.sh |
| settings.json | WARN | Missing PostToolUse hooks |
| settings.local | PASS | Not committed |
| CLAUDE.md | WARN | Missing post-merge protocol |
| Security | PASS | No secrets in git |
After the report, ask the user:
I found N issues (X FAIL, Y WARN). Want me to fix them?
If yes, fix all FAIL items first, then WARN items. For each fix:
After all fixes, re-run the audit to verify everything passes.
npx claudepluginhub damoli1103/claude-code-project-bootstrap --plugin claude-code-project-bootstrapAudits a repository for baseline compliance across 9 categories including code quality, security, CI/CD, testing, and documentation. Emits Markdown report and JSON sidecar.
Bootstraps .claude/ dotclaude config from template if missing, then customizes all files to match project's tech stack, conventions, and patterns. Interactive via user confirmations.
Audits repositories for Claude Code readiness, reviewing CLAUDE.md quality, settings, project organization, and MCP configurations. Provides prioritized improvement suggestions.