render-workflows

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). The URLs are mostly official docs and example code, but they include a raw GitHub-hosted install.sh meant to be piped to sh and a GitHub Releases page for downloading CLI executables — both are distribution vectors for arbitrary code and should be treated as potentially risky unless verified.

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

• Third-party content exposure detected (medium risk: 0.30). In the local development runtime path, an outsider can supply arbitrary JSON inputs via the CLI interactive menu (render workflows tasks list --local → select run → “Provide input as a JSON array”), which the workflow task code will then process inside render workflows dev/the local task server.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs fetching remote example files at runtime to verify API patterns (e.g., https://raw.githubusercontent.com/render-oss/sdk/main/python/example/task/main.py), meaning externally-hosted code/content would be fetched and used to control generated prompts/code.