skillshare

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). These URLs point to a GitHub repo (runkids/skillshare) and raw.githubusercontent.com scripts plus a direct GitHub Releases .tar.gz download — common and transparent distribution channels but they involve direct executable/script downloads (install.ps1, run.sh, and a binary tarball) which are riskier unless you fully trust the repo/author.

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

• Third-party content exposure detected (medium risk: 0.65). Outsider free text can enter the LLM context via skillshare install/update/new when it fetches and parses remote skill repositories (e.g., public GitHub/GitLab pages/README/SKILL.md content) at runtime, which is then used as prompt material for the agent/LLM workflow.

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

• Potentially malicious external URL detected (high risk: 1.00). The included scripts/run.sh performs runtime network fetches and installs/executes a remote binary (it queries https://api.github.com/repos/runkids/skillshare/releases/latest and then downloads/extracts a tarball from https://github.com/runkids/skillshare/releases/download/.../skillshare_${VERSION}_${OS}_${ARCH}.tar.gz via curl), so it executes remote code during runtime.