exec-guardIntelligent command permission classifier — detect dangerous shell commands before execution. Distilled from Claude Code bashClassifier + yoloClassifier.
Install via ClawdBot CLI:
clawdbot install cypress927/exec-guardGrade Limited — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Sends data to undocumented external endpoint (potential exfiltration)
POST → http://localhost:8080/execPotentially destructive shell commands in tool definitions
Exec(Calls external URL not in known-safe list
http://localhost:8080/execAI Analysis
The skill's external HTTP endpoint (localhost:8080) is for local multi-agent sharing, not unauthorized data exfiltration. The primary risk is the inherent danger of arbitrary command execution, which is the skill's documented purpose, not a hidden malicious behavior.
Audited Apr 16, 2026 · audit v1.0
Generated Oct 6, 2026
AI coding agents use exec-guard to safely run build scripts, tests, and package installs without risking memory exhaustion or infinite hangs. The 8KB ring buffer caps output memory while timeout enforcement kills runaway compilers.
CI/CD agents start services in background mode and use watch windows to verify health endpoints before proceeding to the next pipeline stage. Process management endpoints provide logs for failed deployments.
A swarm of market analysis agents shares a central exec-guard HTTP service to run Python models and long-running backtests. Background process handles let agents query intermediate results without blocking each other.
Agents launch GPU training jobs via background execution, poll status and logs periodically, and terminate misconfigured runs early. The shared HTTP mode enables a scheduler agent to monitor all training processes.
Penetration testing agents safely execute scanning tools with strict timeouts to avoid network hangs. Command isolation and process cleanup ensure scans do not leave orphaned processes on target hosts.
Core CLI and single-agent execution remain MIT-licensed while a managed multi-tenant HTTP service with audit logs, RBAC, and process dashboards is offered as a paid subscription. Teams pay per connected agent.
Agent framework vendors embed exec-guard as the default safe execution layer under a commercial redistribution license. OEM licensing includes priority security patches and custom timeout/buffer configurations.
Financial and healthcare enterprises deploy exec-guard behind their firewall with enhanced audit trails, command allowlists, and SIEM integration. Sold as a self-hosted enterprise edition with compliance features.
💬 Integration Tip
Start by sending commands via stdin JSON in CLI mode to validate timeout and buffer behavior, then switch to HTTP server mode when multiple agents need shared process state. Always set timeoutSeconds conservatively and clean up background processes with DELETE /process/:pid to avoid resource leaks.
Scored Oct 6, 2026
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope,...
Security scanner for AI agent skills. 9 built-in detection signatures. Identifies secrets, unsafe execution patterns, and prompt injection. Sub-50ms results.
Wallet anti-theft guard. One-click scan for high-risk wallet approvals to protect user assets. Use when a user asks for a wallet security check, wallet healt...
Comprehensive security audit for an agent's full skill stack. Chains scanner, differ, trust-verifier, and health-monitor into a single assessment with priori...
GEO Audit — AI Search Visibility Checker for ChatGPT, Perplexity, Claude & Gemini. 29-point GEO readiness checklist: robots.txt AI crawler access, Index...
Senior SecOps engineer skill for application security, vulnerability management, compliance verification, and secure development practices. Runs SAST/DAST sc...