cyber-security-consultant

Warn

Audited by Gen Agent Trust Hub on Apr 4, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and install the ACP CLI from a third-party GitHub repository (https://github.com/Virtual-Protocol/openclaw-acp) as part of its setup process.
  • [COMMAND_EXECUTION]: The workflow relies on executing shell commands using the acp CLI tool (acp setup, acp browse, acp job create, and acp job status) to interact with the Cybercentry service.
  • [COMMAND_EXECUTION]: User-supplied data from the query field is interpolated into a shell command string (acp job create ... --requirements '{"query": "..."}'). This pattern is susceptible to command injection or JSON parsing errors if malicious input is provided by an attacker.
  • [COMMAND_EXECUTION]: (Indirect Prompt Injection Surface Analysis) 1. Ingestion points: User-provided query string entered into the job requirements. 2. Boundary markers: No delimiters or explicit instructions to ignore embedded commands are present in the command template. 3. Capability inventory: Execution of system commands via the external acp CLI tool. 4. Sanitization: The documentation advises users to sanitize and escape special characters, but the skill itself lacks any programmatic validation or escaping mechanism.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 4, 2026, 06:22 PM