aris-system-profile
Warn
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a wide array of system-level utilities (e.g.,
strace,perf,nvidia-smi) and run user-specified target processes or scripts. - [COMMAND_EXECUTION]: The instructions explicitly direct the agent to "write and insert profiling code into the target" and "Design the instrumentation based on what you observe in the code". This runtime generation and execution of code within existing user files constitutes dynamic execution and allows for arbitrary code changes to be applied and executed.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it reads and processes external, potentially untrusted source code and command-line arguments. Malicious content within the profiled code could influence the agent's behavior during the instrumentation phase.
- Ingestion points: The agent parses user-provided
$ARGUMENTSand reads the target source code in files specified by the user. - Boundary markers: None are specified to distinguish between the agent's instructions and the content of the data being analyzed.
- Capability inventory: The agent has the ability to modify local files (Step 4), execute shell commands (Step 2/5), and run compiled or interpreted scripts (Step 5).
- Sanitization: No sanitization or validation of the input data or the generated instrumentation code is mentioned.
Audit Metadata