Remote multi-agent workspace CLI with deep Claude integration, native terminal streaming, and self-hosting.
Allowed — no malicious behavior detected
23 low-signal patterns reviewed and cleared.
AI security review
scanned · by lpm-firewall-ai
The inspected lifecycle behavior is limited to local tool extraction and a terminal permission repair. Runtime and agent integration behavior aligns with CLI features, with Claude settings changes requiring an explicit user command; no concrete attack behavior was found.
Static reason
One or more suspicious static signals were detected.; source matched previously finalized malicious package; routed for review; previous stored version diff introduced dangerous source
Trigger
npm postinstall runs the declared extraction and permission scripts; terminal hook setup runs when a user invokes the install-terminal-hooks command.
Impact
These actions support the CLI and terminal mirroring; no unconsented credential transfer or broad agent configuration change was established.
Mechanism
The installer unpacks bundled platform archives, while the optional terminal feature adds the package's own hooks and preserves existing foreign hooks.
Decision evidence
public snapshot
Evidence for AI clean decision
The postinstall runs local tool extraction and a node-pty permission repair; the extraction uses platform archives and writes under the package tools directory.
The Codex session helper reads local session files beneath CODEX_HOME/sessions to build candidate entries.
The assistant bootstrap creates package-owned assistant files only when they are absent.
Claude terminal hooks are installed through an explicit CLI command; the feature preserves foreign hooks and checks for tmux.
The reviewed diagnostics, browser, session-fork, shell, and CLI modules showed no concrete malicious behavior.
Evidence against
The package has a postinstall hook and uses child processes, but the inspected callers use them for local tools and explicit CLI features.
No source-backed credential theft, unconsented broad agent configuration change, or hidden persistence was found.
Behavioral surface
SourceChild process · Crypto · Dynamic require · Environment variables · Filesystem · Network · Shell · Web socket
Supply chainHigh entropy strings · Possible obfuscation · URL strings
ManifestNo manifest risk signals triggered.
Scanned 278 files · 4.87 MB of source
19 external domains referenced
127.0.0.1
accounts.google.com
api.cluster-fluster.com
api.openai.com
auth.openai.com
claude.ai
code.claude.com
console.anthropic.com
developers.google.com
docs.openclaw.ai
github.com
goo.gle
happy.mereith.com
oauth2.googleapis.com
opencode.ai
registry.npmjs.org
veryhappy.dev
www.apple.com
www.googleapis.com
Findings
1 critical · 10 high · 7 medium · 5 low
Automated scan signals; the review checked them and cleared this version. Where they are, most severe file first. Code opens a file's flagged lines, with their evidence, below.