captions-filmcraft

Pass

Audited by Gen Agent Trust Hub on Nov 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill scripts utilize the subprocess module to interact with local binaries and media processing tools.
  • Evidence: fc_captions.py executes filmcraft-cli, ps, ffmpeg, fc-match, magick, fc-list, and fc-cache to perform project modifications and font management.
  • Evidence: fc_inserts.py executes filmcraft-cli to inspect and modify project timelines.
  • [DYNAMIC_EXECUTION]: The skill dynamically modifies the Python module search path to load logic from other local agent extensions.
  • Evidence: fc_captions.py manipulates sys.path to import scripts from the transcribe and filmcraft skills, assuming they are located in standard local directories.
  • [EXTERNAL_DOWNLOADS]: The documentation specifies several third-party tool and library dependencies required for local operation.
  • Evidence: SKILL.md lists prerequisites including filmcraft-cli, ffmpeg, numpy, onnx-asr, and faster-whisper.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes external project data and script files, which defines a potential (though standard) attack surface for prompt injection.
  • Ingestion points: Reads FilmCraft .fcproj files and local .txt script files.
  • Boundary markers: Not present; the content is processed as functional data for the automation.
  • Capability inventory: Executes project-scoped commands via filmcraft-cli and performs file system writes to the project directory.
  • Sanitization: No explicit sanitization of ingested script text is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 12, 2026, 05:01 AM