Skip to content

Security: soovwv/aienvmap

Security

SECURITY.md

Security Policy

aienvmap is designed to be non-invasive by default.

It scans local environment metadata, writes local project files, and does not intentionally upload environment data to a remote service.

Reporting a Vulnerability

Please report security issues through GitHub Security Advisories if available, or open a private contact channel with the maintainer.

Do not include secrets, private environment dumps, access tokens, or production hostnames in public issues.

Operational Safety

  • Normal warnings are advisory and non-blocking.
  • aienvmap doctor --ci is the explicit strict mode for automation.
  • aienvmap should not install, upgrade, downgrade, or remove software by itself.
  • Generated files should be reviewed before committing in sensitive repositories. Treat raw manifests, reconciliation reports, SBOM files, dashboards, and trial portable evidence as local-only by default. Coordination logs can contain operator-supplied text. aienvmap does not edit .gitignore; use a project-specific ignore policy and share only reviewed portable evidence.

Sensitive Data

Current manifests may include local paths, hostnames, shell paths, and runtime versions. Treat generated .aienvmap/ outputs as workspace metadata and decide whether they belong in your repository.

Release provenance

The manual npm workflow requests GitHub OIDC permission and publishes with provenance only from the current main commit when a matching v<version> tag exists. It pins an npm CLI that meets the trusted-publishing floor, rejects already-published versions, and verifies registry version plus integrity afterward. Trusted publisher authentication is the only supported release path and receives no long-lived token. Workflow readiness is not proof that npm-side trusted publishing is configured or that an unpublished release already has provenance.

There aren't any published security advisories