orca-cli

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes examples and commands that place secret values inline (e.g., orca fill --element @e2 --value "s3cret", orca cookie set --value ), which would require the agent to include secrets verbatim in generated commands/outputs.

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

• Third-party content exposure detected (high risk: 0.85). Orca’s browser automation path (orca goto/snapshot/click/fill/exec/eval and related commands) can ingest outsider-authored free text from public web pages at runtime into the LLM context via accessibility-tree snapshots and extracted page text/HTML.