suggest-awesome-github-copilot-agents

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

• Third-party content exposure detected (high risk: 1.00). This skill explicitly fetches and ingests public, user-maintained content from the awesome-copilot repository (e.g., docs/README.agents.md and raw files under raw.githubusercontent.com) using the required fetch/githubRepo steps, and uses that content to decide which agents to suggest, update, or install.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly fetches agent definitions at runtime from the awesome-copilot repository (e.g. https://raw.githubusercontent.com/github/awesome-copilot/main/agents/) and uses those remote agent files—which contain prompt/instruction content—to drive suggestions/updates, so the external content can directly control agent prompts.