Progress Watch
Progress tracking and finish alerts for scripts, jobs and AI agents, no accounts needed
- Pricing
- Open source
- Platforms
- CLI, API, Web
- Runs
- Remotely
- Version
- 0.0.7
- Listing updated
- 1 day ago
Similar MCP servers
About Progress Watch
Progress Watch is an open-source MCP server for developer tools. It runs as a hosted remote server, so there is nothing to install locally.
The docs describe progress tracking for long-running work that nobody is watching, such as crawls, training runs, migrations, CI pipelines and agent tasks. A job creates a task, reports counts as it moves, and closes the task when finished, with a push notification at the end. Counts like 1200 / 50000 pages are shown rather than only percentages, and totals can be revised as a job discovers more work.
Reporting can happen through a CLI, a wrapper command that records start and exit code for commands you cannot modify, a single HTTP request, or an MCP server that lets AI agents create tasks and child steps. Progress is held in memory and expires rather than being stored as history, and a space's UUID serves as its access credential.
The README says the server is AGPL-3.0 licensed and can be self-hosted with Docker, using SQLite by default or PostgreSQL via DATABASE_URL. It suits developers and operators who want a simple status board for background work rather than a full orchestration or logging system.
What it does
- Counted progress reporting
- Tasks show absolute counts such as pages done out of a total, and the total can be updated mid-job.
- Finish notifications
- Web Push alerts fire when a task completes, including when a wrapped command fails.
- Command wrapper
- The run command wraps an existing process, passes its output through, and closes the task with the exit code.
- Agent MCP tools
- An MCP endpoint lets AI agents create a parent task for a job and child tasks for its steps.
- No accounts
- A space UUID acts as the credential, so there is nothing to sign up for or log into.
- Self-hostable server
- The open source server runs as a single container with SQLite or PostgreSQL storage.
Good for
- Watching a nightly crawl or data migration and getting notified when it ends
- Tracking a model training run's progress from a shell script
- Letting an AI agent report steps of a long task to a shared dashboard
- Getting alerts when a CI pipeline job finishes or fails
Pricing
Open source
Open source and self-hostable; a hosted service is also offered.
Prices change. Check progress.watch before you buy.
Progress Watch FAQ
Does it need an account?
No. A space is created and its UUID link is the only credential, so anyone with the link can read and write it.
Is it open source?
The README says the server is distributed under the AGPL-3.0 license and can be self-hosted with Docker.
Is progress history stored?
No. Current state lives in memory and expires, with one latest log line per task, and the docs say there are no charts or history.
Can tasks have subtasks?
Only one level: a task may have steps, but a step may not have its own steps.
Is Progress Watch free?
Yes. Progress Watch is open source, so you can use and self-host it for free.
What are the best alternatives to Progress Watch?
The closest alternatives to Progress Watch are REA, Context7, Chrome DevTools MCP, Codebase Memory and GitHub.
Where this comes from
- Description: written from the project's README, the project's website and its MCP Registry entry.


