launch-run
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute various commands using the 'coval' CLI tool, such as
coval agents listandcoval runs launch. These are legitimate operations for interacting with the Coval evaluation platform. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied input through the
$ARGUMENTSvariable and interactive prompts for identifiers (e.g., agent IDs, persona IDs) to build shell commands. This creates a potential surface for command injection if inputs are not properly sanitized by the agent platform. - Ingestion points: User-provided inputs via
$ARGUMENTSand interactive selection of resource IDs/names. - Boundary markers: None present in the instructions to delimit user input from shell command structures.
- Capability inventory: Shell command execution via the 'coval' CLI utility.
- Sanitization: No explicit sanitization or validation logic is defined within the skill instructions.
Audit Metadata