Add Armorer Guard

[armorer-labs]

Adds Armorer Guard under agent firewalls and runtime protection. It is an open-source local Rust scanner for prompts, tool-call arguments, MCP results, credentials, and exfiltration risk, returning machine-readable reasons suitable for pre-tool-call or post-tool-result enforcement.

[armorer-labs]

Small update for maintainer context: Armorer Guard now has a v0.2.3 release with a refreshed demo and an MCP proxy mode that gates tools/call arguments before execution.

Why this may fit this list: it is a local runtime guardrail for agent boundaries: retrieval ingress, model output, MCP/tool-call args, outbound sends, and memory writes. It returns structured reasons and makes no scanner network calls.

Links:

• Demo: https://huggingface.co/spaces/armorer-labs/armorer-guard-demo
• Release: https://github.com/ArmorerLabs/Armorer-Guard/releases/tag/v0.2.3
• MCP proxy docs: https://github.com/ArmorerLabs/Armorer-Guard/blob/main/examples/mcp_proxy.md

Happy to adjust the placement if there is a more precise category.

[secxena]

Thanks for the submission! Armorer Guard looks on-topic and the repo is actively maintained. Right now the project is still early in terms of adoption, so we are holding off on listing it — happy to revisit as it builds more community traction. Feel free to ping this PR later.