suggest-awesome-github-copilot-prompts

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches and ingests prompt files from the public GitHub awesome-copilot repository (e.g., https://github.com/github/awesome-copilot/blob/main/docs/README.prompts.md and raw files from https://raw.githubusercontent.com/github/awesome-copilot/main/prompts/) via the required #fetch/githubRepo/raw-URL steps, and it uses that untrusted, user-provided content to decide which prompts to suggest, update, and potentially install—allowing third-party instructions to influence agent actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches remote prompt files at runtime from the awesome-copilot repository (e.g., https://raw.githubusercontent.com/github/awesome-copilot/main/prompts/ and https://github.com/github/awesome-copilot/blob/main/docs/README.prompts.md) via #fetch/#runInTerminal, and those fetched files would directly control agent prompts and are required for the skill to operate.