Skip to content

Security: avalonreset/cto-legends

Security

SECURITY.md

Security

Do not report credentials or vulnerabilities in public issues. Use GitHub's private vulnerability reporting on this repository. Never send account passwords, API tokens, or private workspace data with a report.

Supported versions

Version Supported
0.2.x Yes
< 0.2.0 No

Response timeline

  • Initial acknowledgment target: within 72 hours of a private report.
  • Triage and remediation plan target: within 7 days for confirmed security issues in the supported 0.2.x release line.

Security model

  • Closed-vocabulary recipe engine: Catalog entries (cto_legends/catalog.json) are validated against a closed schema (validate_catalog). Remote catalogs never supply arbitrary shell commands, and module repositories are restricted to the avalonreset/ GitHub namespace with mandatory 40-hex commit pins and 64-hex SHA-256 archive checksums.
  • Preview-first operations: install, update, rollback, sync, agent-setup, startup-setup, and isolate-skills preview changes without writing to disk unless --apply is explicitly supplied.
  • Safe archive extraction: Archive extraction rejects path traversal (..), absolute paths, drive prefixes, and symbolic links before touching disk.

There aren't any published security advisories