registry / @negotium/core / 0.1.6

@negotium/core@0.1.6

Bundled into negotium; install negotium instead.

Multi-agent AI runtime — Claude/Codex/Maestro providers, unified event stream, sessions, wiki/skills, local-first storage

AI Security Review

scanned 8h ago · by lpm-firewall-ai

Review flagged AI-agent configuration or capability changes. This remains warn-only unless evidence shows foreign-agent hijack through preinstall/install/postinstall, hidden persistence, exfiltration, remote code execution, or other concrete malicious behavior.

Static reason

No blocking static signals were detected.

Trigger

A host starts an AI turn and an explicit user request switches agents or enables runtime MCP/background-shell capabilities.

Impact

Can alter local AI-agent session state and execute intended agent-runtime capabilities, but no stealth, fixed remote exfiltration, or foreign install-time control-surface mutation was confirmed.

Mechanism

User-gated agent-session bridging plus local runtime service spawning.

Rationale

No concrete malicious chain was found, but the package intentionally modifies local Claude/Codex session control data and exposes shell-capable agent runtime services. Per firewall policy, this is a guarded explicit-user-command AI-agent capability risk that warrants a warning rather than a block.

Evidence

package.jsonsrc/index.tssrc/agents/self-config-core.tssrc/agents/rollout/claude.tssrc/agents/rollout/codex.tsscripts/mcp-patchright-http.mjs~/.claude/projects/<encoded-cwd>/<sessionId>.jsonl${CODEX_HOME:-~/.codex}/sessions/**/rollout-*.jsonlsrc/mcp/background-bash-server.ts

Network endpoints3

@negotium/core@0.1.6

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code