flowus-cli

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFECREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill manages a sensitive API token through the FLOWUS_TOKEN environment variable. The instructions include security best practices, such as redacting credentials from output and avoiding passing tokens via command-line flags to prevent exposure in process lists or shell history.
  • [EXTERNAL_DOWNLOADS]: The skill identifies official installation scripts at https://cdn2.flowus.cn/flowus-cli/install. It explicitly forbids the agent from installing or updating without user approval and requires verification of the binary's SHA-256 hash against official signatures before execution.
  • [COMMAND_EXECUTION]: The skill executes the flowus CLI tool to perform various API operations. It also describes a process for executing downloaded installation scripts (shell or PowerShell) after integrity verification.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface as it retrieves content from FlowUs databases and pages (e.g., flowus markdown get). Because it also possesses write capabilities (e.g., flowus markdown put), there is a theoretical risk of instructions embedded in the retrieved content influencing subsequent agent actions. The risk is mitigated by explicit instructions for write-safety and user confirmation for modifications.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 01:29 PM
Security Audit — agent-trust-hub — flowus-cli