orca-per-workspace-env

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow reads and inlines public provider CLI documentation (or '<cli> --help') during setup (“read that provider's CLI/SDK docs → scaffold…”, i.e., runtime free text from an external web source or CLI help output) into the agent context that is generating the scaffolded scripts.

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

• Potentially malicious external URL detected (high risk: 0.90). The scripts in this skill clone and fetch a repository at runtime (e.g., git clone/fetch using the repoUrl "https://host/org/repo.git") and then build/run code from that repo inside the sandbox/VM, so external content fetched from that URL is executed as part of the workflow.