flow-next-opencode-setup

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The workflow performs setup tasks using shell commands such as mkdir for directory creation, cp for copying binaries and scripts from a local .opencode directory, and chmod +x to grant execution permissions to the installed CLI tool.
  • [COMMAND_EXECUTION]: The skill includes an optional step to star its GitHub repository (gmickel/flow-next-opencode) using the gh api command. This is an interactive process that only executes if the user provides explicit consent through a question prompt.
  • [PROMPT_INJECTION]: The skill creates an indirect prompt injection surface by writing specific operational rules to project documentation files used by other agents. Ingestion points: Reads and writes to CLAUDE.md and AGENTS.md. Boundary markers: Uses <!-- BEGIN FLOW-NEXT --> and <!-- END FLOW-NEXT --> to delimit its instructions. Capability inventory: Explicitly directs agents to use the installed .flow/bin/flowctl binary for state management. Sanitization: Employs static templates without performing complex sanitization of the target file's pre-existing content.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 09:07 AM