Remote multi-agent workspace CLI with deep Claude integration, native terminal streaming, and self-hosting.
Flagged as agent extension risk
Allowed by default with warning: install-time first-party agent extension setup was detected.
AI security review
scanned · by lpm-firewall-ai
Source inspection found a first-party assistant setup that writes agent instructions into its configured assistant home, so the package merits a capability warning. The inspected postinstall performs package tool setup, and the separate Claude settings change requires an explicit command.
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
Running the CLI’s assistant variant triggers the assistant-home bootstrap.
Impact
The assistant session receives package-provided instructions and memory files in its configured home; no confirmed credential transfer, destructive action, or unconsented install-time agent configuration change was identified.
Mechanism
The bootstrap creates a journal directory and writes a package-provided CLAUDE.md and personal memory file when they do not already exist. Terminal mirroring settings require a separate explicit CLI command.
When the assistant variant runs, it bootstraps an assistant home and creates a CLAUDE.md file there if one is absent.
The bootstrap writes the package-provided assistant instructions and personal memory under the configured assistant home; the inspected code does not show it modifying the user’s ordinary Claude settings file.
The separate terminal-hook feature is exposed through an explicit CLI command and writes Claude settings to enable session mirroring.
The manifest’s postinstall runs package tool unpacking and a node-pty permission fix; that hook alone does not establish malicious behavior.
The flagged Codex configuration locations derive the session root from CODEX_HOME or the usual .codex directory.
Evidence against
The code inspected does not show these agent-configuration writes being triggered by npm postinstall.
The terminal mirroring settings change is behind the explicit install-terminal-hooks command and describes mirroring hand-typed Claude sessions in Very Happy Web terminals.
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 276 files · 4.79 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
Where they are, most severe file first. Code opens a file's flagged lines, with their evidence, below.