StyleSeed includes executable CLI scripts, local file reads and writes, plugin manifests, generated artifacts, optional model-exposure paths, and third-party dependencies. If you find a security issue in any of those boundaries, we'd like to hear about it.

Please report privately to bitjaru0402@naver.com rather than opening a public issue. Include:

• A description of the issue and where it is (file/component)
• Steps to reproduce, if applicable
• The version / commit you found it on

We aim to acknowledge reports within a few days and will credit you (if you want) once a fix ships.

Supported versions are published tags or published artifacts, plus separately identified security fixes. main is not an immutable release by itself.

In scope:

• executable CLI and local automation under engine/.claude/skills/** and scripts/**
• local files written or read by manifests, generated bundles, evidence files, and learning records
• plugin/package boundaries, including plugin manifests and MCP exposure contracts
• model-exposure boundaries where a local package may become visible to a connected client/model
• path, symlink, traversal, containment, dependency, and supply-chain issues in maintained distributions

Out of scope:

• third-party deployments that add their own infrastructure, secrets, or wrappers
• downstream projects that copied or modified the engine and no longer match a published artifact
• unverifiable claims about a local install without a reproducible artifact or exact file evidence

• The implemented default/core install contains no learning MCP.
• Known high-risk identity patterns are blocked in local learning; this is a guardrail, not an anonymization guarantee.
• A prepared learning package is local and untransmitted unless a future optional bridge is separately implemented and verified.