Skip to content

Security: kschlt/cookframe

SECURITY.md

Security policy

Cookframe converts recipe sources you point it at into a durable, provider-independent form. It fetches remote URLs and calls a model provider, so its security posture is part of the project, not an afterthought.

Reporting a vulnerability

Please report suspected vulnerabilities privately through GitHub's private vulnerability reporting on this repository (Security → Report a vulnerability). Do not open a public issue for a vulnerability.

We aim to acknowledge a report within a few days and to keep you updated as we work on a fix. Please give us reasonable time to remediate before any public disclosure.

Secrets

  • No real secret is ever committed. .env.example carries placeholders only; real values live in your environment or a secret store and are loaded at run time.
  • CI runs a secret scan on every push and pull request. A finding fails the build.
  • If you believe a secret was committed, rotate it immediately and report it as above — rotation matters more than removing it from history, because anything pushed should be assumed captured.

Fetching remote content

Cookframe fetches source URLs on the server side. Any code that fetches a URL on behalf of a caller must refuse internal targets — loopback, private-network, and link-local addresses (SSRF protection). The url-fetch-security CI job exists to enforce this invariant as the fetch layer is built.

Supply chain

The repository is public and open to contribution. Dependencies are kept minimal, pinned through the lockfile, and installed with npm ci in CI and in the container. Dependency update automation is .github/dependabot.yml, in the tree rather than a setting, so it is versioned and reviewable.

The protections that live on GitHub itself — branch protection, secret scanning — cannot be read back by anything in this repository, so they are not claimed here. What is known about each one, who checked it and when, is recorded in docs/repository-protections.md, including the ones that are still off.

There aren't any published security advisories