create-paperclip-bundled-skill

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

• Third-party content exposure detected (medium risk: 0.30). The required workflow only ingests outsider-authored text when capturing tweet/X sources via xc get and xc search (Steps 0 in SKILL.md), and otherwise reads only builder-provided files or pinned first-party metadata at build/test time.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill’s external-reference flow pins and fetches an upstream GitHub repo at build/runtime (see the gh command "gh api repos/mvanhorn/last30days-skill/commits/v3.3.0 --jq .sha"), and the pinned files list includes agents/openai.yaml and multiple scripts which can directly control prompts or execute code.