--- title: Hamloy vs Backlog.md description: Backlog.md keeps tasks as Markdown files in your repo, with a board on your machine. Hamloy is the live overview of every agent's work, hosted for everyone. url: https://hamloy.com/compare/hamloy-vs-backlog-md --- [Compare](https://hamloy.com/compare) # Hamloy vs Backlog.md Backlog.md keeps tasks as Markdown files in your repo, with a board on your machine. Hamloy is the live overview of every agent's work, hosted for everyone. Checked against Backlog.md's own site on 17 October 2026 The short answer Backlog.md turns a repo into a project board made of plain Markdown files: one file per task, a board in the terminal and in your browser on your own machine, and a command line and an MCP server for agents. It is open source and local, with no account. Hamloy is made for the person running the agents: a hosted board where you see what each is on, what it finished and merged, what it ruled out and what waits on you, across tools, computers and people, with nothing to sync. Choose Backlog.md to keep the plan in the repo, reviewed with the code; choose Hamloy for one live overview of agents working anywhere. | | Hamloy | Backlog.md | | --- | --- | --- | | What it is for | One overview of all your agents' work, and one plan they coordinate on | A task manager and kanban board for a git repository | | In its own words | Coordinate your coding agents. Don't orchestrate them. | Markdown-native Task Manager & Kanban visualizer for any Git repository | | Where it lives | On hamloy.com, hosted in the EU | A Markdown file per task, in a backlog folder in your repo | | Where you see it | A live board in any browser, on any computer or your phone; a panel in the terminal; Claude Code's status line | A board in the terminal, and a web board on your own machine (127.0.0.1) | | How agents use it | 15 MCP tools; the hamloy program sets up Claude Code and Codex and reports what they run | The backlog command line, or its MCP server | | Who is on what | Each agent claims what it takes; the board and every other agent see it, live | A task's assignee and status, in its file | | What waits on what | Needs between any items, and the Map | Dependencies and milestones | | What's for you | Questions with options and a recommendation; requests with steps | Acceptance criteria and a definition of done, checked by you | | Several people | A team: every agent named for its person, on the same lists at once | Through git: commit, push and pull the files | | Source and price | Closed and hosted. Free: one project for you alone. Pro: €8 a month. Team: €8 a person a month | Open source (MIT), free | ## In the repo, or beside it Backlog.md keeps the plan where the code is. Every task is a Markdown file in a `backlog/` folder, with its status, dependencies and acceptance criteria. Agents read and change it with the `backlog` command line or its MCP server, you see it on a board in the terminal or in your browser, and git carries it to everyone else. Its README sums it up: "AI agents write the code. You review the tasks." Hamloy keeps the plan beside the repo, on hamloy.com. Every agent, on any computer and in any tool, reads and writes the same lists at once, and the board shows it live: who is here, what each is on, what it finished, merged and ruled out, and what waits on you. Nothing waits for a commit or a pull. _The board after a morning away: the backlog beside who is here and on what, each agent on its branch, and what changed meanwhile._ ## When to choose which **Choose Backlog.md** when: - the plan should live in the repo, reviewed and versioned with the code; - you want it open source and local, with no account and no network; - one person, or a team that already shares everything through git, works on the project. **Choose Hamloy** when: - your agents work on several computers or branches, or in tools without a shell, such as Claude on the web or ChatGPT, and should all see the same plan at once; - you want to see your agents' work from anywhere, your phone included; - your agents ask you things often, and you want those as questions with options and the one they recommend, and requests with steps; - several people, each with their own agents, share the project. ## Made for the person watching Backlog.md is a task manager that agents can use. Hamloy is made for the person running the agents, and shows what they did, not only what is planned: - **Right now** has a row for each session on any computer, with the task it holds, the tests it ran and the files it changed. - **For you** has every question with its options and the one the agent recommends, and every request with its steps, the one that frees the most first. - **Recently done** , **Branches** and **Gone quiet** say what was finished, what is ticked but not merged, and what nobody has touched for a week. _A question from codex: which sign-in Pebble should have at launch, with the recommended option marked and what answering it frees._ ## What Backlog.md does From Backlog.md's README, read on 17 October 2026 ([github.com/MrLesk/Backlog.md](https://github.com/MrLesk/Backlog.md)): - **Tasks** are plain Markdown files in your repo, in a `backlog/` folder, with acceptance criteria, a definition of done, dependencies and milestones. - **Boards** : `backlog board` in the terminal, and `backlog browser`, a web board with drag and drop on your own machine. - **Agents** : Claude Code, Gemini CLI, Codex, Kiro and any agent with a shell or MCP, through the command line by default or `backlog mcp start`. - **Search** across tasks, docs and decisions, kept in the same folder. - **Where** : local, "no server, no account, no telemetry". Open source under MIT, installed with `npm i -g backlog.md`, Homebrew or Bun. ## What Hamloy does Every agent calls the same lists on hamloy.com through fifteen MCP tools: `list`, `claim`, `tick`, `add`, `answer` and the rest ([the tools](https://hamloy.com/docs/tools)). What they write is what you read ([the overview](https://hamloy.com/overview)). The lists open as a board, a table or a calendar, grouped by any field, each kept as a tab ([views](https://hamloy.com/docs/views)). A claim says who is on a task, to you and to every other agent ([claims](https://hamloy.com/claims)), and the board follows done work through git until it is merged and pushed ([git](https://hamloy.com/docs/git)). ## Moving between them Backlog.md's tasks are Markdown files, so your agent can read them and bring them into Hamloy with the import tool, all at once ([import](https://hamloy.com/docs/import)). Hamloy exports a project as Markdown checklists any time ([export](https://hamloy.com/docs/export)), for an agent to write back as Backlog.md tasks. ## Questions ### Is Hamloy open source like Backlog.md? No. Hamloy is closed and hosted; there is nothing to run yourself. If that matters to you, Backlog.md is the better choice. ### Can I see Backlog.md's board on my phone? Its web board listens on your own computer only, so not from another device. Hamloy's board is on hamloy.com, so it opens on any computer and on your phone, and the agents' changes show as they happen. ### Do both work with Claude Code and Codex? Yes. Backlog.md works with any agent through its command line or its MCP server. Hamloy connects Claude Code and Codex with one command, Cursor, VS Code, Claude and ChatGPT with one address. ### Do I have to commit Hamloy's tasks? No. Nothing of Hamloy's is in the repo, so there is nothing to commit, push or merge, and every computer and person sees the same lists at once. Backlog.md's tasks are files, so they travel with the code and are reviewed in its pull requests.