Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Openclaw Pixel Agents Dashboard
v1.0.0Real-time pixel art ops dashboard for OpenClaw deployments. Visualizes agent activity as character sprites in a shared office with live activity bubbles, har...
⭐ 0· 328·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description match the code and instructions: the server tails local JSONL session files, serves a React canvas, shows hardware stats, can call the OpenClaw gateway (/tools/invoke) to spawn agents, and can optionally sync remote agents over SSH. Required resources (gateway token, agentsDir, optional SSH creds) are appropriate for those features and nothing in the repo requests unrelated cloud/provider credentials or system-wide secrets.
Instruction Scope
SKILL.md instructs a normal npm-based setup (npm install, run setup wizard or build/start server). The runtime behavior described in SKILL.md (auto-discover agents, generate config, tail JSONL session logs, call gateway API, optionally rsync remote agents) is implemented in the included server code. I did not find instructions that ask the agent to read unrelated system config, exfiltrate secrets, or contact unexpected third-party endpoints beyond the gateway/update checker; the setup wizard and config loader legitimately access config paths (~/.openclaw/agents, ~/.config/pixel-agents) and environment variables for tokens.
Install Mechanism
No opaque download/install mechanism is present: the project is intended to be installed/started via npm (package.json) and runs local Node/tsx code. Dependencies come from npm (package-lock.json) rather than arbitrary remote archives. Assets are bundled under public/assets. No suspicious extract-from-URL or unknown-host downloads were observed in the provided files.
Credentials
The repo does not require env vars in registry metadata, but the config system supports ${ENV_VAR} expansion and the code will use OPENCLAW_GATEWAY_TOKEN (or gateway.token in the config) and optional remote SSH credentials (password or keyPath). These are proportionate to the dashboard's needs (gateway auth and remote agent syncing). Caution: remoteAgents supports plaintext passwords and recommends sshpass — storing plaintext SSH passwords in config is risky. Prefer SSH key auth or environment variables and review who can read the config files.
Persistence & Privilege
The skill does not request always:true and is user-invocable. It does not appear to modify other skills or global agent configurations. It runs as a normal user-mode Node process and uses per-user config locations; nothing indicates elevated/system-wide privilege is required.
Assessment
This package is internally consistent with its stated purpose, but review a few things before running it: 1) Inspect skill/scripts/setup.sh (it auto-discovers agents and writes config) so you know what filesystem paths it will access and what it will write. 2) Use environment variables or SSH key auth for gateway and remote access rather than putting passwords in dashboard.config.json. 3) The dashboard tails local JSONL session files (by default under ~/.openclaw/agents) — run it only on systems where you expect that data to be available and safe to display. 4) The 'ham radio' update-checker will contact update endpoints — if you need strict network controls, run the server in an isolated environment or block outbound requests. 5) As with any npm bundle, consider running in a dedicated user account or container the first time you try it, and audit network traffic if you have concerns about exterior communication.bin/pixel-agents.cjs:24
Shell command execution detected (child_process).
server/hardware.ts:69
Shell command execution detected (child_process).
server/services.ts:82
Shell command execution detected (child_process).
server/version.ts:29
Shell command execution detected (child_process).
Patterns worth reviewing
These patterns may indicate risky behavior. Check the VirusTotal and OpenClaw results above for context-aware analysis before installing.Like a lobster shell, security has layers — review code before you run it.
latestvk97fr0t38pcfzhdss2yv9wkq7182qp8e
License
MIT-0
Free to use, modify, and redistribute. No attribution required.