The Agent Skills Directory

[SAFE]: The skill implements a read-only diagnostic workflow. All tools listed in the frontmatter (mcp__goofish__auth_status, mcp__goofish__search_items, mcp__goofish__item_view, mcp__goofish__item_get) are designed for data retrieval and status checking without modification capabilities.
[SAFE]: The skill contains explicit anti-patterns that discourage malicious marketplace behaviors such as data brushing, spamming, or account-linking, promoting safe seller practices instead.
[SAFE]: There are no signs of obfuscation, hardcoded credentials, or unauthorized network operations. The diagnostic logic is transparently documented in the provided reference files.
[SAFE]: While the skill processes external marketplace data (which is a surface for indirect prompt injection), the risk is minimal as the skill lacks execution or write capabilities that could be exploited via injected content. The mandatory evidence chain for this surface includes: Ingestion points in search_items and item_view outputs; no explicit boundary markers or sanitization are defined, but the capability inventory is strictly limited to read-only diagnostic reporting.

goofish-shop-diagnosis