sync-opencode
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The orchestrator script (
scripts/run-opencode-sync.mjs) uses the Node.jsspawnmethod to execute a sequence of local scripts, includingsync-config.mjs,sync-skills.mjs,sync-hooks.mjs, andsync-agents.mjs. - [DYNAMIC_EXECUTION]: The skill's primary function involves generating an executable JavaScript bridge plugin (
.opencode/plugins/easy-claude-hooks.js) from local settings. Additionally, the orchestrator dynamically discovers and executes test files using the Node.js test runner (--test). - [INDIRECT_PROMPT_INJECTION]: The skill manages a data processing surface where project-level configuration and agent definitions are ingested and transformed into generated commands and mirrors.
- Ingestion points: Processes
opencode.json,.claude/settings.json, and.claude/agents/*.mdfrom the project environment. - Boundary markers: No specific delimiters or "ignore instructions" markers are documented for the generated command outputs.
- Capability inventory: The skill possesses extensive file system modification capabilities and the ability to spawn subprocesses via the Node.js runtime.
- Sanitization: The skill documentation notes that descriptions in generated commands are YAML-escaped to prevent formatting breakages.
Audit Metadata