OpenAFW is a local agent firewall — keep your secrets off the model, the API relay, and the supply chain. Local credential masking, per-route model routing, and security detectors on the wire.
SOCKS5 and DNS proxy for greywall
What can your AI agent do to the real world? Scan your code. See which tool calls have zero checks
Tool-neutral attack corpus for AI agent egress security
Migrated → github.com/zistica/korveo (rebranded). Archived.
What can your TypeScript AI agent do to the real world? Scan your code. See which tool calls have zero checks
Python verifier for Pipelock action receipts (Ed25519-signed, chain-linked)
Personal AI agent firewall skill for blocking prompt injection, unsafe tool calls, data exfiltration, malicious skills, and memory poisoning.
Policy-based firewall and honeypot middleware for AI agents accessing websites
GitHub Action for running agent-driven CI jobs through Pipelock with verifiable egress evidence.
Clampd - Runtime Security for AI Agents. Dev infrastructure and orchestration.
Provide an open test corpus to assess and improve AI agent egress security through validated cases and automated workflows.
Add a description, image, and links to the agent-firewall topic page so that developers can more easily learn about it.
To associate your repository with the agent-firewall topic, visit your repo's landing page and select "manage topics."