unquantized-brainstorm
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads .mp3 audio files and JSON metadata from the author's public GitHub repository (luchtan-GIT/m4r-data) via raw.githubusercontent.com and GitHub Release assets. These resources are vendor-owned and necessary for the skill's creative purpose.
- [COMMAND_EXECUTION]: The m4r_ingest.py script executes local system audio players such as afplay, ffplay, mpg123, paplay, or aplay using subprocess.run. The commands are safely constructed using argument lists, which prevents shell injection.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external JSON metadata and audio files which could serve as a surface for indirect prompt injection.
- Ingestion points: Data enters the agent's context when it reads the metadata file /tmp/m4r_brainstorm_meta.json and the audio file /tmp/m4r_brainstorm.mp3 downloaded via the fetch command.
- Boundary markers: The instructions do not define explicit delimiters or include warnings to ignore instructions embedded in the external content.
- Capability inventory: The skill allows the agent to execute shell commands via the ingestion script and read/write local files.
- Sanitization: There is no evidence of validation or sanitization performed on the fetched JSON content before it is processed by the agent.
Audit Metadata