Infraveil Agent Guard

io.github.infraveilhqPublished by GitHub account infraveilhqv0.1.1Updated Oct 24, 2026

Gate an AI agent's destructive actions behind human approval, with a tamper-evident log.

VerifiedSTDIODesktop onlySecurity & Monitoring

Overview

AI-generated overview

Gates an AI agent's destructive actions behind human approval and keeps a tamper-evident log of them.

What you can do
Infraveil Agent Guard sits between an AI agent and the actions it wants to take, holding dangerous operations for human approval before they run. It records actions in a ledger and an approval queue, and can operate in enforce mode, where risky actions are blocked pending approval, or audit mode, where they are only logged. A severity threshold decides which actions are gated.
When to use it
Worth adding when an assistant is allowed to run commands or change systems and you want a human checkpoint before anything destructive happens. It suits teams that need an auditable record of what an agent attempted and who approved it.
Requirements
Runs as a local process on the user's machine, installed from the PyPI package infraveil-guard over stdio; desktop only. Configuration is by environment variable: INFRAVEIL_GUARD_HOME for the ledger and approval queue directory, INFRAVEIL_GUARD_MODE for enforce or audit, and INFRAVEIL_GUARD_THRESHOLD for the gated severity level. No authentication is declared.
Before you install
In enforce mode the server blocks actions pending approval, so an agent can stall until someone responds; audit mode never blocks, so it only observes. The ledger and approval queue are written to the directory named by INFRAVEIL_GUARD_HOME, which may hold sensitive records of agent activity. Review the threshold setting before relying on it, since actions below it are not gated.

Installation

In SourceWeft

  1. Open Infraveil Agent Guard in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

This repository has no README.

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v0.1.1LatestSep 16, 2026

Score

25% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts

    Declare a GitHub repository with a README, or serve resource_documentation in the OAuth protected-resource metadata.

  • License10 pts

    Declare a license in the repository or the registry entry.