autosync-ic-skills

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow fetches public web content from https://skills.internetcomputer.org/.well-known/skills/index.json and then downloads skill files from https://skills.internetcomputer.org/.well-known/skills/<skill>/<file>; those outsider-authored texts/files are written into .claude/skills/ and become LLM-readable when Claude discovers/loads the synced SKILL.md content.

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

• Potentially malicious external URL detected (high risk: 1.00). This skill installer downloads and registers a SessionStart hook that fetches and runs remote content from https://skills.internetcomputer.org (notably the installer script at https://skills.internetcomputer.org/.well-known/skills/autosync-ic-skills/scripts/sync-ic-skills.sh and the runtime index at https://skills.internetcomputer.org/.well-known/skills/index.json), and that fetched content is used at runtime to populate .claude/skills (which directly controls agent prompts/skills and can include executable skill code), so it is a runtime external dependency that can control prompts/execute code.