RuleDrift

io.github.ri7inPublished by GitHub account ri7inv0.1.1Updated Oct 24, 2026

Test whether AI agents retain critical instructions as conversations grow.

VerifiedSTDIODesktop onlyDeveloper ToolsAI & ML

Overview

AI-generated overview

RuleDrift tests whether an AI agent still follows critical instructions as a conversation grows long, exposing results as optional MCP tools for AI coding and M

What you can do
Run rule-drift tests against an agent to find the first conversation point where a rule fails, using local Ollama models or a bring-your-own agent setup. Produce terminal, JSON and JUnit reports and stable exit codes for CI. Expose these checks to AI coding clients through optional MCP tools. Initialize a project configuration before running tests.
When to use it
Useful when you change a prompt, model, memory system or agent workflow and want to check that important instructions still hold over long or distracting conversations, or when you want drift checks wired into CI.
Requirements
Runs locally as a Node.js package over stdio; requires Node.js 22+ and Ollama, with a model pulled locally (for example qwen3:1.7b). No paid model API, account, backend or telemetry is required. The MCP mode is started with a project root path.
Before you install
It runs locally and does not sit in front of production requests, so it is a test harness rather than a runtime guard. It writes project configuration and report files, and MCP mode is given access to a project root path; review that path before use.

Installation

In SourceWeft

  1. Open RuleDrift 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

[RuleDrift logo]

RuleDrift

Catch the moment your agent drifts from its rules.

RuleDrift tests whether an AI agent still follows important instructions after a conversation becomes long or distracting.

Run it like a test suite when you change a prompt, model, memory system, or agent workflow. It runs locally and does not sit in front of production requests.

Quick start

Requires Node.js 22+ and Ollama.

bash
ollama pull qwen3:1.7bnpx rule-drift initnpx rule-drift test

No paid model API, account, backend, or telemetry is required.

What you get

  • The first conversation point where a rule fails
  • Terminal, JSON, and JUnit reports
  • Local Ollama and bring-your-own-agent support
  • Optional MCP tools for AI coding clients
  • Stable exit codes for CI

MCP

bash
npx -y rule-drift mcp --root /path/to/your/project

See the MCP setup guide for client configuration and safety details.

Documentation

Contributions are welcome—see CONTRIBUTING.md. Licensed under MIT.

Source: README.md at commit 5fb43b4

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

50% 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
  • License10 pts