agentguard

io.github.agentwaresPublished by GitHub account agentwaresv0.1.9Updated Oct 25, 2026

MCP policy proxy: spend caps, approvals for destructive tools, kill switch, dry-run, audit log.

VerifiedSTDIODesktop onlyDeveloper Tools Security & Monitoring

Server details

Identifier: io.github.agentwares/agentguard
Transport: STDIO
Runtime: Desktop only
Trust: Verified
Published: Sep 16, 2026

Security note

MCP servers receive tool arguments from a conversation and may perform external actions. Review what a server can do before you enable it, and give it only the access it needs.

The registry verified the GitHub account agentwares for this listing. That does not show the account is the product's maker.

SourceWeft's checks cover whether the listing is complete and the endpoint answered; they do not review the server's code or what its tools do.

Add a badge to your README

Show readers this MCP server is listed on SourceWeft. The badge links to this page.

[![Listed on SourceWeft](https://sourceweft.com/badge/mcp/io.github.agentwares%2Fagentguard.svg)](https://sourceweft.com/mcp/io.github.agentwares%2Fagentguard?utm_source=readme&utm_medium=badge)

Overview

AI-generated overview

A local MCP policy proxy that sits in front of tool calls to enforce spend caps, approvals for destructive tools, dry-run mode, a kill switch and an audit log.

What you can do: Acts as a policy proxy over MCP tool calls, applying spend caps and requiring approvals before destructive tools run, with a dry-run mode and an audit log of activity. Exposes its own read-only tools, including a status tool, a report tool and a view tool, when no config file is present. Can halt every tool call through a kill switch.
When to use it: Worth adding when an assistant runs tools that cost money or can change or delete data, and you want limits, approval gates and an audit trail around those calls.
Requirements: Runs locally as a stdio process via npx from the npm package. Optionally reads a config file whose path is given by AGENTGUARD_CONFIG, defaulting to ./agentguard.yaml; without it, only the server's own read-only tools are served. Desktop only.

Before you install

The kill switch is controlled by the AGENTGUARD_KILL environment variable; setting it to 1 halts every tool call. The proxy mediates tool calls and keeps an audit log, so review what it records. Approvals and spend caps only apply to calls routed through it.

Installation

In SourceWeft

Open agentguard in the dashboard and add it to a workspace.
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

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

Version history

v0.1.9LatestOct 8, 2026
v0.1.7Oct 7, 2026
v0.1.6Oct 7, 2026
v0.1.3Sep 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.