HN Showcase
Back to browse
GitHub Repository

The open-source agent firewall. Prevent AI agents from leaking data, using dangerous tools, and importing poisoned dependencies.

41 starsJavaScript

ClawMoat – Open-source host-level security for AI agents

by ildar·Feb 25, 2026·2 points·0 comments
Visit ProjectView on HN

AI Analysis

●●●BangerSolve My ProblemWizardry

Real threat model for agentic systems: detects prompt injection pre-execution.

Strengths
  • •Directly solves emerging class of vulnerabilities (agent prompt injection)
  • •Zero dependencies, sub-millisecond scanning—no ML model overhead
  • •YAML policy engine lets teams define custom rules per agent capability
Weaknesses
  • •Early stage (6 stars); detection evasion arms race likely unfolds over time
  • •Coverage depends on constantly updating threat patterns as attacks evolve
Category
Security
Target Audience

Engineers building AI agents with LangChain, CrewAI, AutoGen, or OpenAI Agents

Similar To

Rebuff.ai · Vigil by Mushroom Garden

Similar Projects

AI/ML●●Solid

Circuit Breaker – runtime cost ceilings for AI agents

Simple wrapper preventing runaway agent costs before they hit your credit card.

Solve My ProblemNiche Gem
joakimOslo
101mo ago