maestro-update
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute Node.js scripts (maestro-install.js and maestro-render-orchestrator.cjs) using environment variables for path resolution. These scripts perform file synchronization and content rendering tasks within the project and plugin directories.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect injection surface because it programmatically updates the SKILL.md file by ingesting data from .claude/maestro.json. 1. Ingestion points: .claude/maestro.json (read during the rendering process). 2. Boundary markers: The skill utilizes specific HTML comment tags (e.g., , ) to isolate generated content from user-modified prose in the orchestrator skill file. 3. Capability inventory: The skill utilizes shell commands via node to execute scripts that have the ability to read and overwrite files in the .claude directory. 4. Sanitization: The skill relies on internal script logic to process the JSON configuration before rendering it into the markdown skill body.
Audit Metadata