upskill

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). All links point to GitHub/raw.githubusercontent content (common and generally legitimate) but they reference third‑party/individual repos and include a remote installer fetched via raw.githubusercontent (curl ... | bash) — a high‑risk pattern unless you inspect and trust the scripts and maintainers.

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

• Third-party content exposure detected (high risk: 0.85). At runtime, the required workflow ingests outsider-authored free text from the user’s own Claude Code session logs (e.g., [USER] messages and session.log captured by the SessionEnd/UserPromptSubmit hooks) into the Teacher/Student build pipeline, which then feeds that text into the LLM context for skill generation/validation.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill directs users to run a remote installer that fetches and executes code (curl -sSL https://raw.githubusercontent.com/HKUDS/Upskill/main/cc-integration/install.sh | bash and the equivalent git clone https://github.com/HKUDS/UpSkill && cd UpSkill/cc-integration && bash install.sh), which clearly executes remote code as part of the install/runtime path and is required for the tool to function.