opencode-plugin-creator
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation of plugins that ingest untrusted data from user messages and tool outputs, creating a potential surface for indirect prompt injection. Ingestion points: Data enters the plugin context via hooks such as
chat.messageandtool.execute.before(documented inreferences/hooks.md). Boundary markers: While the template does not use specific delimiters, the skill provides logic for blocking suspicious calls. Capability inventory: Plugins can execute shell commands using Bun's shell API and perform filesystem operations (SKILL.md,references/hooks.md). Sanitization: The skill includes a pre-built security utility (assets/template/src/plugin/guard.ts) to validate and block access to sensitive file paths. - [SAFE]: The skill is a development toolset that follows legitimate patterns for the OpenCode ecosystem and explicitly includes defensive coding examples.
Audit Metadata