The latest published 0.x minor on npm receives security fixes. Earlier minors do not. create-starter is pre-1.0; minor bumps may include breaking changes alongside fixes.
| Version | Supported |
|---|---|
0.4.x |
✅ Latest |
< 0.4 |
❌ |
Please do not file public GitHub issues for security problems. Use one of the private channels below.
- GitHub private advisory (preferred) — open a draft at https://github.com/starter-series/create-starter/security/advisories/new. This routes directly to the maintainer and triggers GHSA workflows.
- Email —
wantcongz@gmail.comwith subject[create-starter security] <short description>.
Please include:
- The affected version (or commit SHA).
- A minimal reproduction or proof-of-concept.
- Your assessment of impact (e.g. RCE, path traversal, SSRF, supply-chain).
- Whether you intend to publish your own write-up; coordinated disclosure timelines are negotiable.
- Acknowledgement within 72 hours.
- Triage + initial assessment within 7 days.
- Fix or mitigation within 30 days for High/Critical severity; 90 days for Moderate/Low.
- Public advisory published via GHSA once a fix is available, with credit to the reporter unless they prefer anonymity.
In scope:
- The
@starter-series/createnpm package (CLI + MCP server). - The
.mcpbClaude Desktop bundle. - The
create-starterClaude Code plugin and bundledcreateskill. - The
audit_release/audit_cd/audit_securityMCP tools. - The GitHub Actions workflows in
.github/workflows/.
Out of scope:
- Downstream projects scaffolded from a template — those are independent repos; report vulnerabilities to the respective template repo.
- Issues in third-party MCP clients (Claude Desktop, Cursor, Windsurf) — report to the client vendor.
- Vulnerabilities that require already-compromised credentials or full filesystem access (these are post-exploit conditions).
This repo passes its own audit_security 8/8 HARDENED — gitleaks (SHA-pinned), CodeQL (advanced workflow), npm audit, license check, --ignore-scripts on every install, Dependabot grouped updates, native GitHub secret-scanning + push protection, and the anthropics/claude-code-security-review Action (SHA-pinned).