filmcraft-cli-timeline
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads the required video editing engine and desktop application from the author's GitHub repositories. It implements rigorous integrity checks by verifying every downloaded file against hardcoded SHA-256 hashes defined in the runtime and desktop lock files.
- [COMMAND_EXECUTION]: The editing engine and auxiliary tools (such as hdiutil and codesign) are executed as subprocesses. On macOS, these processes are wrapped in a strict sandbox using
/usr/bin/sandbox-exec, which limits their access to user-authorized directories and prevents unauthorized network operations during editing. - [DYNAMIC_EXECUTION]: Uses the Model Context Protocol (MCP) to manage the lifecycle of the editing engine. Communication is handled through a structured JSON protocol with validation against a local command catalog to ensure only supported operations are performed.
- [INDIRECT_PROMPT_INJECTION]: The skill processes structured editing plans and media metadata. It mitigates potential risks from malicious data by enforcing strict JSON parsing rules, including key uniqueness and type validation, and by using internal permission gates before executing any native commands.
Audit Metadata