pinto is a new open-source tool that brings Scrum backlog management to the terminal, storing every item as a plain-text Markdown file with TOML frontmatter. This means teams can track changes with git diff, branch the backlog alongside code, and review updates in pull requests—just like source code. It is local-first, requiring no server or account, and works with AI agents without depending on them.
Why pinto Exists
The creators of pinto argue that popular project management tools like Jira, Asana, and Notion have become bloated over time. Creating a ticket often involves navigating required fields and workflow settings; running a Sprint requires configuring permissions, automations, or dashboards. Even a simple status change can feel costly when the tool is slow to open. Scrum is meant to be a lightweight framework for rapid inspection and adaptation, but the tooling can start to serve the process instead of the team. pinto aims to keep the team's focus on the Product Backlog, the Sprint, and the work flowing across the board—hence the name, derived from the Japanese word for “focus.”
Your Backlog Belongs in the Repository
Running `pinto init` creates a `.pinto/` directory next to your code. Each Product Backlog Item (PBI) is a Markdown file with TOML frontmatter containing an ID, title, status, rank, labels, and timestamps. Because the backlog is made of ordinary files, your existing Git workflow applies: review changes with `git diff`, branch the backlog with the code that changes it, discuss updates in pull requests, and recover history with `git log`. pinto also supports linking commits to PBIs, maintaining a shared Definition of Done, and recording dependencies—all without a separate service. Git is not mandatory; the default file backend works without it, but the data format can participate naturally in a Git workflow whenever the project does.
Scrum Workflow in 90 Seconds
A typical workflow starts by initializing the board and adding PBIs with points and labels. The `pinto list` command shows a concise table of items with their status, title, points, and labels. Rank controls the Product Backlog order, and priority changes are as direct as `pinto reorder T-4 --top`. To plan a Sprint, you create a new sprint with a goal, start and end dates, then add the top-ranked PBIs and start the sprint. The `move` command shifts items through statuses like "in-progress" and "done," and `pinto board` displays a Kanban view. Teams can set WIP limits in a config file or use `pinto kanban` for an interactive terminal UI where they can drag cards, edit them, and manage dependencies.
During the Sprint, `pinto sprint burndown` renders a burndown chart directly in the terminal, showing remaining points against the ideal trend. Before committing, teams can estimate capacity with `pinto sprint capacity`. At the end of the Sprint, closing it and running `pinto sprint velocity` and `pinto cycletime --sprint S-1` reveals velocity, cycle time, and lead time. The entire board can be created in under a minute, and the timing values in the example are measured in seconds—though a real project would show distributions across days and build useful velocity trends over successive Sprints.
Safe Automation with or Without AI
Most read-only commands support `--json` output. For writes, pinto provides `p automate`, a deliberately narrow boundary for agent-driven changes. It accepts a structured JSON plan made of argument arrays. Plans contain validated argv, never shell code, so there is no shell-injection path. The `--dry-run` flag executes the plan against an isolated copy of the board without modifying the real one, and `--json` reports every command as valid, succeeded, failed, or skipped, along with created and updated item IDs. This allows AI agents to read the backlog, propose a plan for a Sprint, and show a dry run before applying anything—keeping the human in the loop.
Installation and Availability
pinto is written in Rust and can be installed from crates.io with `cargo install pinto-cli`. Building from source requires Rust 1.89 or later. The repository README provides full instructions. The project is open-source and available on GitHub under the MIT license.