bmo is a local-first command-line issue tracker backed by a single SQLite file, designed for use by both human developers and AI agents operating in a terminal. It requires no server, no network dependency, and no external services. Issues are identified by BMO-N IDs.
For an example of agents and skills that have been instructed to use bmo check out the bmo-agent-setup project (no need to run the program: the agents and skills are plain markdown files).
bmo was inspired by and adapted from docket, an issue tracker for AI agents written by ALT-F4-LLC.
In addition, the code in this repository was written by Claude Code. The repo owner directed this work.
From crates.io (recommended):
cargo install bmoFrom source:
cargo install --path .Pre-built binaries:
Download the latest binary for your platform from GitHub Releases.
bmo init
bmo create --title "First issue" --priority medium --kind task
bmo list
bmo board
bmo webIf you use pi-code, the pi-bmo extension integrates bmo natively
into pi. This extension adds typed bmo_* tools that the LLM calls directly, in addition to a live board status widget, board
state injected into every agent turn's context, and /board /next /plan slash commands.
pi install git:github.com/erewok/pi-bmoA companion pi-code extension, bmo-dev-team, adds a
full software development team of agents (staff-engineer, project-manager, senior-engineer,
qa-engineer, ux-designer) that plans and executes work through bmo.
pi install git:github.com/erewok/bmo-agent-setupSee docs/pi-extension.md for the full integration guide.
- docs/commands.md — Complete command reference
- docs/agents.md — Integration guide for AI agents
- docs/pi-extension.md — pi-code extension and dev-team suite
- docs/web.md — Web interface
- docs/data.md — Data storage and database location
- docs/migration-from-docket.md — Migrating from docket
- docs/development.md — Building and running tests
Apache-2.0. See LICENSE file.
bmo was inspired by docket by ALT-F4-LLC. Please credit that project for the ideas in this project.