negotium@0.2.15
Install the Negotium multi-agent runtime and CLI with one package
AI Security Review
scanned 2mo ago · by lpm-firewall-aiNo concrete credential exfiltration, stealth persistence, or unconsented install-time AI-agent control-surface write was found. The package nevertheless has significant user-invoked agent lifecycle and shell capability, warranting a warning rather than a block.
Decision evidence
public snapshot- package.json postinstall downloads two executable helpers.
- install-*.mjs write verified binaries under ~/.negotium.
- Runtime can create synthetic ~/.claude rollout files for agent switching.
- Fixture JSONL embeds inert AI/developer instruction transcripts.
- Installers use fixed GitHub release URLs and hard-coded SHA-256 checks.
- No install hook writes Codex/Claude/MCP configuration.
- Remote bridge code only permits configured loopback URLs.
- Source network calls are provider, configured service, or localhost runtime traffic.
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 child process execution.
dist/hosted-agent.jsView on unpkg · L8Findings
Affected versions and remediation
This report applies to negotium@0.2.15.
See all versions of negotium for other recorded verdicts.
- Review the evidence and your use of negotium@0.2.15 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.