opencode-bridge

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The skill documentation describes the use of the opencode CLI to execute prompts and retrieve JSON responses. This is a functional requirement of the bridge and does not involve unauthorized command execution.\n- [PROMPT_INJECTION] (LOW): The skill is subject to indirect prompt injection (Category 8) because it reads and processes message data from an external database that may contain content from untrusted sources.\n
  • Ingestion points: Message history is read from ~/.opencode/opencode.db (messages table).\n
  • Boundary markers: None present; the bridge processes raw database records.\n
  • Capability inventory: The skill can execute CLI commands via opencode -p and interact with an MCP server (openwork-mcp-bridge).\n
  • Sanitization: There is no documented sanitization or validation of the message content before it is used in subsequent operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:05 PM