Skip to content

Security: luz0Dm/ide-super-worker

Security

SECURITY.md

Security Policy

Supported Versions

The public repository tracks the latest release only until a formal release process is established.

Reporting A Vulnerability

Open a private security advisory on GitHub, or contact the maintainer through the repository issue tracker and request a private disclosure channel. Do not post exploit details or live credentials in public issues.

Operational Notes

  • Never commit .env, real provider keys, GitHub tokens, terminal logs, or generated archives.
  • Rotate any credential that was pasted into chat, shell history, issue text, or logs.
  • Keep SANDBOX_ROOT narrow. The worker refuses paths outside this root.
  • Prefer scoped_patch for write tasks so out-of-scope edits are marked failed.
  • Keep ALLOW_BYPASS_PERMISSIONS disabled unless you fully trust the caller.
  • Use HTTPS provider endpoints outside private networks.
  • Treat third-party model gateways as data processors for the prompts and files you send through start, a configured semantic gate, or the legacy worker-LLM lane.

Built-In Guardrails

  • Secret redaction for common API key variables and bearer tokens.
  • Realpath-based sandbox checks.
  • Optional scoped patch enforcement.
  • Deterministic search, read_pack, diff_digest, and shell evidence paths; analysis, judgment, and code review remain in the main model by default.
  • Public evidence routing is limited to deterministic tools such as read_pack, diff_digest, and get_artifact_slice; historical quality/fan-out modules have no public MCP entry point.
  • Semantic verification is a separate post-execution gate and does not enable general worker analysis or review.
  • Optional git worktree isolation for parallel jobs.

There aren't any published security advisories