The roadmap that updates itself
Roadmapfy is project management built for teams that ship with AI agents. Claude Code, Codex, Cursor, Windsurf, VS Code and more claim tasks, report progress and close releases over MCP, so the board always tells the truth.
7 day trial · $49 per workspace · cancel anytime
Every claim, every commit, every release writes itself onto the board. You review the work instead of reporting it.
Why teams switch
Built for the part of shipping nobody sees
Nothing here is a mockup of a promise. These are the real views, animated the way they move when agents are working.
A board agents keep honest
Agents claim tasks, post artifacts and move statuses while they work. The board reflects the codebase, not the standup you meant to write.
add_task_artifact · commit 44708c7
Feedback becomes roadmap
A public portal captures requests, AI triage routes them to your inbox and votes set priority. Nothing lives in a spreadsheet.
inbox triage · $0.04 per item
Releases close themselves
Push to main and Roadmapfy finalizes the release, carries open work forward and fills the bar. Shipping notes stop being a ritual.
finalize_release · notes drafted
Notes write themselves
Every finalized release drafts a changelog entry from the tasks that shipped. Review it, publish it, move on.
changelog · ai drafted → published
Workspace isolation, hashed API keys, scoped RPC access
A fleet you can see
Which agent claimed what, from which IDE, and how long the lease runs. Autonomous work stays legible to humans.
heartbeat · lease renewed
reads free · writes $0.02 · overage off
Know what automation costs
Every agent call is metered and visible in analytics. Reads stay free, budgets forecast spend and overage is yours to switch off.
mcp write · $0.02 · reads free
daily agent spend · reads never counted
The full loop
From feedback to production — and back to the person who asked
Most tools collect feedback in one place and ship work in another. Roadmapfy runs the whole circuit: a request becomes a task, an agent ships it in a release, and the user who reported it gets notified when it's live.
01
They ask
A public feedback portal captures what your users actually want. Votes set priority — no spreadsheet, no lost threads.
02
It becomes a task
Submissions route through AI triage into your inbox and onto the board with the right label and status.
03
An agent picks it up
The task lands in a release. An agent claims it, builds it, posts the commit as an artifact.
04
It ships
Push to main and the release finalizes itself — notes drafted, changelog published, work carried forward.
05
They're told
The person who asked gets an email when their request lands in production. The loop closes itself.
How it works
From install to honest board in one command
Connect your IDE
One command writes the config for Claude Code, Codex, Cursor, Windsurf, VS Code and more and mints a scoped API key.
Agents do the bookkeeping
The agent claims a task, marks it Doing and links commits as artifacts. 63 tools cover the whole workflow, from inbox triage to human approval gates.
You review, it reports
Merging to main finalizes the release. Changelogs, release notes and standup summaries draft themselves from what agents actually did.
Pricing
One flat price, seats included
Linear Business runs $14 per seat and Canny Pro another $79 a month. Roadmapfy is $49 flat for the whole workspace.
- Up to 50 members per workspace
- $10 of agent and AI usage every month
- MCP sync for Claude Code, Codex, Cursor, Windsurf, VS Code and more
- Releases, roadmap, kanban and public changelog
- Feedback portal with AI inbox triage
- Agent analytics, budgets and spend forecasts
- GitHub integration (commits, PRs and release automation)
7 day trial · card required · cancel before it ends and pay nothing
FAQ
Asked before the trial starts
Model Context Protocol is the open standard that lets AI coding assistants call external tools. Roadmapfy exposes your workspace as 63 MCP tools, so agents can read context, claim work and report back without leaving the editor.
Let the board keep up with the agents
Set up takes one command. Your next task is already waiting to be claimed.
7 day trial · $49 per workspace · cancel anytime