mcloud-projects

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Executes mcloud CLI commands and jq for project listing, retrieval, and deletion. These operations are scoped within the allowed-tools configuration to prevent arbitrary command execution beyond the mcloud and jq toolsets.- [PROMPT_INJECTION]: Identifies a surface for indirect prompt injection where untrusted data from project names or handles could influence agent behavior.
  • Ingestion points: Data ingested from mcloud projects list and mcloud projects get output as described in SKILL.md.
  • Boundary markers: Absent; the instructions do not define delimiters for external project data or include specific instructions to ignore embedded commands.
  • Capability inventory: The agent is authorized to execute mcloud and jq commands, which could be targeted if malicious strings are present in project metadata.
  • Sanitization: No specific sanitization or filtering is applied to CLI outputs before they are processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 01:14 PM
Security Audit — agent-trust-hub — mcloud-projects