skillfence
Runtime security monitor for OpenClaw skills. Watches what your installed skills actually DO — network calls, file access, credential reads, process activity. Not a scanner. A watchdog.
Why use this skill?
Protect your OpenClaw environment with SkillFence, the runtime watchdog. Detect malicious network calls, file access, and credential theft in real-time.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/deeqyaqub1-cmd/skillfenceWhat This Skill Does
SkillFence is a specialized runtime security monitor for your OpenClaw environment, acting as an vigilant watchdog for your installed skills. Unlike static scanners that analyze code before installation, SkillFence operates in real-time to monitor the actual behavior of your skills while they are executing. It tracks critical system activities including network calls, file system access, credential reading, and process spawning. If a skill attempts to open a reverse shell, exfiltrate data, or access sensitive configuration files without authorization, SkillFence detects it immediately. It is an essential component for any OpenClaw user concerned about supply-chain attacks or hidden vulnerabilities in third-party extensions.
Installation
To integrate SkillFence into your OpenClaw ecosystem, execute the following command in your terminal:
clawhub install openclaw/skills/skills/deeqyaqub1-cmd/skillfence
Once installed, you can trigger the monitor by running: node {baseDir}/monitor.js --watch
Use Cases
SkillFence is designed for both proactive and reactive security management. Use it when vetting a new skill before you grant it full permissions, or as a continuous background service to monitor your current fleet of extensions. It is particularly effective for post-incident audits; if your system exhibits strange behavior, running --audit-log provides a clear trail of what files were accessed and which network connections were made by active skills. It provides peace of mind when installing community-developed skills that may have unknown dependencies or obfuscated logic.
Example Prompts
- "OpenClaw, run a full system scan with SkillFence to check for any malicious network activity or unauthorized credential access."
- "I just installed a new third-party skill. Use SkillFence to perform a deep scan on that specific skill and tell me if it is safe to keep."
- "SkillFence just alerted me to a medium-severity issue. Can you show me the audit logs for the last hour to help me identify which process triggered this?"
Tips & Limitations
SkillFence is not a replacement for code reviews; it is a behavioral monitor. While it excels at detecting runtime threats like C2 connections or unauthorized file reads, it cannot predict every possible malicious outcome. For maximum security, run periodic system scans using the --scan command. Note that intensive monitoring may have a negligible impact on system performance during heavy computational loads, so schedule full scans during idle periods if you are running resource-constrained hardware.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-deeqyaqub1-cmd-skillfence": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: network-access, file-read, code-execution
Related Skills
HyperStack — Agent Provenance Graph for Verifiable AI
The Agent Provenance Graph for AI agents — the only memory layer where agents can prove what they knew, trace why they knew it, and coordinate without an LLM in the loop. Timestamped facts. Auditable decisions. Deterministic trust. Ask 'what blocks deploy?' → exact typed answer. Git-style branching. Three memory surfaces: working/semantic/episodic. Decision replay with hindsight bias detection. Conflict detection. Staleness cascade. Utility-weighted edges that self-improve from agent feedback. Agent identity + trust scoring. Time-travel to any past graph state. Works in Cursor, Claude Desktop, LangGraph, any MCP client. Self-hostable. $0 per operation at any scale.
zero-rules
Intercept deterministic tasks (math, time, currency, files, scheduling) BEFORE they hit the LLM. Saves 50-70% on token costs by resolving simple queries locally with zero API calls.