FlowCheck FAQs

Question 1

What is FlowCheck?

Accepted Answer

FlowCheck is a Git hygiene safety layer designed for AI-first development. It monitors Git repositories in real-time and nudges AI agents to maintain clean, manageable commit histories, preventing large, hard-to-review changesets.

Question 2

How does FlowCheck help AI agents maintain clean commit history?

Accepted Answer

It monitors the Git state, tracking metrics like uncommitted lines and time since last commit. When configurable thresholds are exceeded, FlowCheck provides recommendations and nudges directly to AI agents via clear tool interfaces to make timely checkpoint commits.

Question 3

What 'flow health' metrics does FlowCheck track?

Accepted Answer

FlowCheck tracks crucial metrics for code hygiene, including minutes since the last commit, the number of uncommitted lines, and the count of uncommitted files. It then provides a status (ok, warning, danger) based on these metrics.

Question 4

Is FlowCheck compatible with existing AI coding assistants?

Accepted Answer

Yes, FlowCheck is specifically built for agentic coding workflows and uses the Model Context Protocol (MCP). It can be integrated with tools like Claude and Cursor by adding its rules and configuring it as an MCP server, enabling direct communication with AI agents.

Question 5

Can I customize the monitoring thresholds in FlowCheck?

Accepted Answer

Absolutely. FlowCheck uses a configuration file (`~/.flowcheck/config.json`) where you can dynamically adjust thresholds such as `max_minutes_without_commit` and `max_lines_uncommitted` to suit your project's specific needs and preferences.

FlowCheck

FlowCheck

主な機能

ユースケース

主な機能

ユースケース