alterlab-genai-soundtrack-composer
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill follows an 'Autonomous Agent' pattern focused on film scoring and music production. It uses legitimate external tools (Suno) and provides structured templates for music composition tasks.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that involves reading untrusted external data (scripts, project files, video edits) to maintain creative continuity.
- Ingestion points: SKILL.md (Workflow step 1 and Agentic Protocol section) instructs the agent to read scripts, video edits, and existing project files.
- Boundary markers: None identified; the agent is not explicitly told to ignore instructions embedded in those project files.
- Capability inventory: The skill has the capability to read files, write files to the project directory, and search the web for platform updates.
- Sanitization: No explicit sanitization or validation of the input file content is described.
- Assessment: While this creates a surface for indirect prompt injection (where a malicious script file could try to influence the agent's scoring decisions), the risk is minimal and tied to the core functionality of the skill.
Audit Metadata