oh-my-opencode@5.1.48
⚠ Under reviewThe Best AI Agent Harness - Batteries-Included OpenCode Plugin with Multi-Model Orchestration, Parallel Background Agents, and Crafted LSP/AST Tools
AI Security Review
scanned 10d ago · by lpm-firewall-aiThe scanner’s dynamic-evaluation lead resolves to a test helper, and the inspected postinstall behavior is plugin cache invalidation plus platform checks. The selected optional platform binary is an executable dependency whose source is unavailable here, so the review remains unresolved without evidence that it is malicious.
Decision evidence
public snapshot- The manifest declares platform-specific optional binaries pinned to version 5.1.48; the launcher resolves the platform candidate and executes it.
- The JavaScript launcher selects a resolved binary and passes it the CLI arguments; the binary’s source was not available for this review.
- The postinstall hook runs a version/platform check and invalidates cache directories whose names match this plugin.
- The flagged dynamic evaluation occurs in a test helper that builds a selector function from a test source string.
- No concrete attack behavior was established in the inspected JavaScript or lifecycle hook.
- The scanner-highlighted evaluation is in a test file, and the launcher’s platform selection is consistent with the package’s declared binary layout.
Source & flagged code
23 flaggedSource creates an unconsented AI-agent control surface through install-time mutation or a default unauthenticated remote skill channel.
package.jsonView on unpkg · L1Package defines install-time lifecycle scripts.
package.jsonView on unpkgFindings
Affected versions and remediation
This report applies to oh-my-opencode@5.1.48.
See all versions of oh-my-opencode for other recorded verdicts.
- Review the evidence and your use of oh-my-opencode@5.1.48 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.