@octavus/agent@1.0.15
Octavus Agent CLI - run one of your cloud Octavus Agents on a machine you control, using that machine as the agent's computer. The agent lives in the cloud (fully configured and observable); the local machine only supplies the computer and the trigger.
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThe package is not malicious by the install-hook standard, but it provides a dangerous remote AI-agent capability with local command execution and remote communication. Warn so consumers can assess the intended trust boundary.
Decision evidence
public snapshot- The user-invoked CLI provisions shell, filesystem, browser, and computer-control tools for a cloud agent.
- Its code MCP accepts a command and sends it to a shell execution helper, so remote agent instructions can produce local command output.
- The package connects runs to the Octavus platform and stores an API key locally.
- package.json has no preinstall, install, postinstall, or other lifecycle script.
- The privileged behavior is reached only by running the octoagent command; it is not install-time or import-time.
- No source evidence shows covert credential harvesting, stealth persistence, or destructive automatic behavior.
Source & flagged code
13 flagged · loading sourcePackage source references child process execution.
dist/computer-mcp/dist/index.jsView on unpkg · L53Source file is highly similar to a previously finalized malicious package; route for source-aware review.
dist/computer-mcp/dist/index.jsView on unpkgFindings
Affected versions and remediation
This report applies to @octavus/agent@1.0.15.
See all versions of @octavus/agent for other recorded verdicts.
- Review the evidence and your use of @octavus/agent@1.0.15 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.