Skip to content

Security: ilovepixelart/toro

SECURITY.md

Security Policy

Supported Versions

Security fixes are released only for the latest 1.x version. Releases before 1.0.0 are no longer maintained; upgrade to the latest 1.x release to receive a fix. The supported Python and Redis versions are listed in docs/versioning.md.

Version Supported
1.x ✅
0.x ❌

Reporting a Vulnerability

Please do not open a public GitHub issue for security problems.

Report vulnerabilities privately via GitHub's private vulnerability reporting form. This routes the report directly to the maintainer through a private advisory and keeps the details out of the public issue tracker until a fix is available.

When reporting, please include:

  • The affected version(s) of toro-queue, plus the Python, redis-py and Redis server versions
  • A minimal reproduction (producer and worker code, job options, observed vs expected behavior)
  • The impact you believe the issue has (code execution, data integrity, cross-queue access, denial of service, etc.)

Response Expectations

  • Acknowledgement: within 7 days of the report.
  • Triage and fix window: targeted within 30 days for confirmed issues, depending on severity and complexity.
  • Disclosure: coordinated via the GitHub advisory. A CVE will be requested where applicable, and a patched release will be published to PyPI before the advisory is made public.

Scope

In scope:

  • The toro package source in this repository.
  • The published distribution on PyPI (toro-queue).

Out of scope:

  • Vulnerabilities in redis-py or the Redis server itself: report those to redis-py or Redis.
  • Access that follows from reaching the Redis server. Whoever can reach the Redis a queue uses can do anything that queue can do; securing that Redis is the operator's part of the model.

Security model

docs/security.md describes what toro guarantees (JSON-only serialization, no dynamic dispatch from Redis, no string-built commands, key-safe identifiers), what the operator owns, and the limits worth knowing.

There aren't any published security advisories