aidaw
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local command-line interface (CLI) binary named
aidawto interact with the DAW application. It provides specific logic for locating the binary on different operating systems (macOS, Windows, Linux) by reading configuration files at fixed paths or using system tools likemdfindandopen -b. - [DATA_EXPOSURE]: Several tools (
import_audio,export_wav,master_project) facilitate reading from and writing to the local filesystem. These tools accept file path parameters, allowing the agent to access audio assets on the user's machine. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from the DAW project and installed plugins, creating an attack surface where maliciously crafted project metadata or plugin information could attempt to influence agent behavior.
- Ingestion points:
get_project(reads project JSON),list_plugins(reads plugin metadata). - Boundary markers: None mentioned; the agent is instructed to read the data to "see the current situation."
- Capability inventory: The skill can execute shell commands via the
aidawCLI, read local files viaimport_audio, and write local files viaexport_wav. - Sanitization: No explicit sanitization or validation of the ingested project data is described.
Audit Metadata