skills-sh-search

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (low risk: 0.10). The workflow searches public skill registries (skills.sh and agentskills.io) and ingests descriptions and metadata of outsider-authored packages, falling under low risk since the agent must actively search/fetch specific items.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to install external packages via npx (npx --yes skills find and npx --yes @agentskill.sh/cli search) from third-party repositories without strong first-party provenance or immutable identifiers, constituting an unverifiable operational runtime dependency.