Security fixes are applied to the latest 1.0.x release line. Older releases do not
receive backports; upgrade to the most recent version on PyPI (iterabledata).
| Version | Supported |
|---|---|
| 1.0.x (latest) | Yes |
| < 1.0 | No |
Please report vulnerabilities privately through GitHub Security Advisories: open the repository's Security tab and choose Report a vulnerability. Do not open public issues for unpatched vulnerabilities.
You can expect an acknowledgement within 7 days. We will coordinate a fix and a release before public disclosure, and credit reporters in the release notes unless you ask otherwise.
This library intentionally reads and writes files in many formats, including files from untrusted sources. Relevant attack surface includes:
- Format parsers and compression codecs handling malformed or hostile input
(see the
input-securitycapability spec underopenspec/specs/for the hardening requirements already adopted). - SQL identifier handling in the database ingest backends
(
iterable/ingest/), which routes all identifiers throughiterable/ingest/identifiers.pyand binds row values as parameters. - Optional AI/LLM integrations (
iterable/ai/) — never send untrusted file contents to a cloud provider without review; use the built-in redaction helpers where offered.
Bugs in these areas are in scope for this policy. Vulnerabilities in optional third-party dependencies should also be reported upstream to the dependency.