rewelo 0.7.2 · experimental

Prioritise your backlog by value for effort

rewelo ranks tickets with the relative weight method: score each one for benefit, penalty, estimate and risk, and it tells you what returns the most for the least. From the command line, or from your AI assistant through MCP.

Read the docs View on GitHub

Command line

npm install -g rewelo
rw project create Acme
rw ticket create --project Acme --title "Login page" \
  --benefit 8 --penalty 5 --estimate 3 --risk 2
rw ticket list --project Acme --sort priority

Every command is in the CLI reference.

MCP server in Docker

docker run --rm -i --init --cap-drop=ALL \
  --read-only --memory=256m \
  -v rw-data:/data ghcr.io/sebs/rewelo serve

Add it to Claude Desktop or Claude Code as described in the MCP guide.

The examples walk through importing, scoring, tracking and reporting on a backlog, each as a Claude Code prompt and as rw commands.

How the method works

Value = Benefit + Penalty

Cost = Estimate + Risk

Priority = Value / Cost

Every dimension uses the agile Fibonacci scale: 1, 2, 3, 5, 8, 13, 21. Weights let you emphasise what matters to you, and relative weights compare a ticket with the whole backlog or with a tagged subset. The calculations page has the details.

Tags, not fields

prefix:value tags such as state:wip or feature:auth organise tickets, one value per prefix, with every change in an audit log.

Reports

Summary, backlog health, score distribution, lead and cycle times, an event stream, diffs since any date, and a self-contained HTML dashboard.

Your data, one file

Everything lives in a single SQLite file via Node's built-in node:sqlite. Export and import CSV or JSON, including full history.

Built for assistants

MCP tools let Claude create, score, tag, relate and report on tickets, with the same validation as the CLI.