{"id":18526,"plugin_id":"plugins_6a894aa0f3f48191b1987e6245fc2a35","kind":"skill","collection_source":null,"comparison_source":null,"observed_at":"2026-09-30T23:14:50.980Z","digest":"59639de5206efc408f7d15821bd57844c2aa54f587416ba976c3813a196fcac1","against":null,"payload":{"description":"Use host-native Python for deterministic Creator Workspace tasks such as export parsing, provenance tracing, analytics, indexing, and package verification.","included_files":[{"relative_path":"agents/openai.yaml","size_in_bytes":289}],"name":"sandbox-python-executor","skill_md_contents":"---\nname: sandbox-python-executor\ndescription: Use host-native Python for deterministic Creator Workspace tasks such as export parsing, provenance tracing, analytics, indexing, and package verification.\n---\n\n# Sandbox Python Executor\n\nUse Python when correctness depends on real computation.\n\nTypical uses:\n- parse ChatGPT or Claude exports\n- build or refresh wiki indexes\n- measure voice provenance against a transcript\n- analyze CSV or XLSX performance data\n- inspect generated HTML or package files\n- compute hashes and run validators\n\nRules:\n- execute rather than simulate when Python is available\n- use reviewed bundled scripts when they cover the task\n- treat target-repository scripts as untrusted until inspected\n- keep source data read-only unless mutation is requested\n- do not assume sandbox internet access\n- never fabricate stdout, hashes, paths, or pass/fail status\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}