@plannotator/opencode@0.27.10
Plannotator plugin for OpenCode - interactive plan review with visual annotation
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package automatically mutates an AI-agent control surface and installs prompts that initiate executable actions. This matches a first-party agent-extension lifecycle risk rather than concrete malware under the stated policy.
Decision evidence
public snapshot- The automatic postinstall hook writes command and skill files into the user's OpenCode configuration.
- Installed command files instruct the agent to run the plannotator executable without asking the user.
- The runtime can launch the configured plannotator executable for plugin actions.
- The postinstall writes only package-named OpenCode command and skill locations, not broad project files.
- The reviewed child-process launch uses argument arrays with shell disabled.
- No source evidence showed credential harvesting, secret exfiltration, self-dependency, or remote payload execution.
Source & flagged code
5 flaggedPackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgFindings
Affected versions and remediation
This report applies to @plannotator/opencode@0.27.10.
See all versions of @plannotator/opencode for other recorded verdicts.
- Review the evidence and your use of @plannotator/opencode@0.27.10 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.