Skip to content

Security: fmadore/iwac-ai-pipelines

Security

.github/SECURITY.md

Security Policy

Reporting a vulnerability

Report security issues privately through GitHub's private vulnerability reporting rather than as a public issue. Expect a reply within two weeks.

Scope

These pipelines hold two kinds of credential, both read from a local .env that is git-ignored and never committed:

  • AI provider API keys — Gemini, OpenAI, Mistral, OpenRouter.
  • Omeka S API keysOMEKA_KEY_IDENTITY and OMEKA_KEY_CREDENTIAL, which grant write access to a live archive.

Reports about credentials leaking into logs, terminal output, saved payload dumps, or checkpoint files are in scope, as are ways to make a write script PATCH the archive without passing through common/write_guard.py.

What these pipelines send to third parties

Whole archival documents are sent to third-party model providers. OpenRouter requests are pinned to data_collection: "deny" so they are routed only to backends that do not retain data; the Gemini, OpenAI, and Mistral APIs are governed by those vendors' own terms. Anyone reusing this code on their own collection should confirm that arrangement suits their material before running anything at scale.

There aren't any published security advisories