@rallycry/conveyor-agent@11.0.48
Conveyor Agent Runner v10 - PTY harness for the task chat (SDK harness for audit/project-chat). Agent-as-User architecture with BaseService patterns. Works locally too.
AI Security Review
scanned 12d ago · by lpm-firewall-aiThe inspected runtime boot code silently replaces Claude configuration and rewires credential storage, which warrants a warning for agent configuration capability. No npm install attack or concrete exfiltration chain was established.
Decision evidence
public snapshot- The manifest has no install hook; its prepublishOnly script runs the build.
- Runtime boot code deletes and recreates the user home Claude configuration file, seeds project trust and model settings, and wires Claude credentials into pod-local storage.
- The agent harness configures broad tool permissions and command hooks when the harness runs.
- The grimoire setup uses a supplied GitHub token only when the workspace submodule is missing.
- The inspected configuration changes are in runtime boot code, not an npm install lifecycle.
- The inspected sources show no credential transmission to an unrelated endpoint or other concrete attack chain.
Source & flagged code
7 flagged · loading sourceSource writes persistence or remote-access backdoor material.
dist/boot-KDC4OLVL.jsView on unpkg · L35A package entrypoint or install-time lifecycle script reaches a source file with blocking dangerous behavior.
dist/boot-KDC4OLVL.jsView on unpkg · L35Findings
Affected versions and remediation
This report applies to @rallycry/conveyor-agent@11.0.48.
See all versions of @rallycry/conveyor-agent for other recorded verdicts.
- Review the evidence and your use of @rallycry/conveyor-agent@11.0.48 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.