mastering-engineer

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes Bash in its allowed-tools list. However, the instructions explicitly discourage running manual shell scripts, directing the agent to use specific, managed MCP tools instead.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external data, including user-defined configuration files (mastering-presets.yaml) and audio file metadata. While this represents a potential attack surface for indirect prompt injection, the risk is mitigated by the skill's narrow focus on technical audio processing parameters.
  • Ingestion points: mastering-presets.yaml via load_override, and album metadata via find_album.
  • Boundary markers: Not explicitly defined for YAML or metadata parsing.
  • Capability inventory: File read/write, specialized audio processing tools (master_audio, analyze_audio), and bash access.
  • Sanitization: Not explicitly documented in the instructions.
  • [SAFE]: The skill employs path resolution via a dedicated tool (resolve_path), which helps ensure that file operations remain within the intended application directory structure (~/bitwize-music/), adhering to the principle of least privilege.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 03:13 AM