$ loading_
Monitor agent actions, gate money-moving steps, and verify security verdicts independently.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "com.fidacy/ai-agent-firewall" MCP server from askskill: Run: claude mcp add 'com-fidacy-ai-agent-firewall' -- npx -y @fidacy/mcp
Developers or platform teams can use it when deploying AI agents that act on external systems, monitoring each step and adding allow-or-block controls for sensitive operations.
When an agent may trigger payments, transfers, or similar money-moving actions, teams can review them with this tool before execution continues.
When teams need to audit or review an agent's security judgment, it can provide independently verifiable verdicts to support traceability and accountability.
It is an MCP firewall-style tool for AI agents. The provided description says it watches agent actions, gates money-moving operations, and provides independently verifiable verdicts.
From the description, it watches all agent actions but especially gates money-moving operations. For the exact rule scope, see the source repository.
No documentation excerpt was provided, so the installation steps, runtime environment, and key requirements cannot be confirmed here. Please see the source repository.
Add DeFi wallet safety, token risk checks, and transaction simulation for AI agents.
Protect AI agents with scans for injections, leaks, risky URLs, and passwords.
Add offline, deterministic pre-action safety gates with signed verdicts for AI agents.
Enable secure agent payments, API key vaulting, and governed spending limits.
Add human approval, policy enforcement, and audit logs to AI tool calls.
Create verifiable delegation chains for AI agents with scoped multi-hop authorization