swe-pro-agents@3.0.1
25 OpenCode agent profiles (22 subagents + 3 primary) + 14 skills (caveman, skill-creator, teach-me, documentation, svg-hero-generator, humanizer-pro, flowchart-html, high-quality-flowcharts, next-best-thing, nextreme-optimization, opencode-skill-creator,
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is a first-party AI-agent extension installed automatically into a global host control surface. It is suspicious lifecycle risk rather than proven malware because no concrete exfiltration, remote payload, or self-reinstall chain was found.
Decision evidence
public snapshot- Automatic post-install hook runs the installer.
- Non-interactive installs default to all agents, skills, background tools, and an enabled goal system.
- Installer copies a package-controlled plugin into OpenCode’s global plugin directory.
- Installed plugin registers hooks and can prompt OpenCode sessions.
- No runtime dependency on this package name is declared.
- No install-time network request, credential harvesting, or payload download was found.
- The plugin is package-named and its session prompting is gated by its goal feature.
Source & flagged code
8 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgPackage source references dynamic require/import behavior.
bin/swe-pro-agents.jsView on unpkg · L29Findings
Affected versions and remediation
This report applies to swe-pro-agents@3.0.1.
See all versions of swe-pro-agents for other recorded verdicts.
- Review the evidence and your use of swe-pro-agents@3.0.1 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.