Lightpanda

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). These URLs include direct downloads of unsigned "nightly" binaries hosted on a relatively unknown GitHub repo (lightpanda-io/browser) — a common vector for distributing executables — so despite the script attempting checksum verification via the repo metadata, installing/running these untrusted nightly binaries poses a moderate-to-high risk.

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow for this skill is using Lightpanda to goto/fetch arbitrary user-supplied URLs and then ingest the resulting page content (e.g., markdown, semantic_tree, mcp://page/markdown) into the agent’s LLM context, which is outsider-authored free text from public web pages.

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

• Potentially malicious external URL detected (high risk: 0.90). The install script (scripts/install.sh) fetches and installs a remote binary from GitHub (e.g., https://github.com/lightpanda-io/browser/releases/download/nightly/lightpanda-x86_64-linux and the GitHub API URL https://api.github.com/repos/lightpanda-io/browser/releases/tags/nightly), verifies it, makes it executable, and runs it for testing, so the skill runtime/install step fetches and executes remote code.