Blume
Blume · Coding · Productivity
Blume is a local-first developer tool for watching and improving coding agents. It keeps every agent you run, from Codex to Claude Code to Cursor, in one narrow, always-on window, so you can see what's working, what's waiting on your approval, and what it costs. Think of it as a coding agent tracker for the whole fleet. Rules, skills, hooks and corrections pile up in every coding agent setup. Blume maps the whole graph and keeps it maintained, and your conversation history stays on your own device.
About Blume
What Is Blume
Blume is a lightweight desktop companion for people who run AI coding agents all day. Ever lost track of which terminal was still working? That's the problem it solves. Instead of juggling separate terminals and dashboards for Codex, Claude Code, and Cursor, you get one persistent panel that shows each agent's state: finished, working, or blocked on your approval. That last one matters. When an agent needs a decision, you see it immediately rather than discovering it twenty minutes later.
The deeper idea is context hygiene. Every coding agent setup accumulates its own rules, skills, hooks, and one-off corrections, and those silently drift out of sync with what you actually want. Blume tracks those hidden files and rules. It also watches for mismatches between your agent setup and your chat instructions, and shows you the exact diff before anything changes.
We should be clear about the limits. Blume is a monitor, not an agent. It won't write your code or fix a broken build for you. Several of its more ambitious features, such as auto-fixes, analytics, and a shared team domain model, are marked "Soon" or "Next" on the official site. So the current build is narrower than the long-term vision. It runs as a desktop app, and the local-first design means history lives on your machine. That's great for privacy, but it means no cloud sync across machines.
Getting Started
- Go to blume.codes and download the desktop app for your system.
- Install it and open the narrow always-on window, which parks on your desktop without hogging space.
- Connect the coding agents you already use, like Codex or Claude Code, and follow the short setup guide for each provider.
- For usage tracking, complete the provider login or permission step so Blume can read your local session data.
- Start a normal agent session and watch its status, token burn, and any rule conflicts surface in the panel.
Product Information
A quick look at Blume's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- Developers running more than one coding agent
- Teams that share agent rules
- Privacy-conscious engineers
Tasks
- Watching agent status
- Tracking plan limits and token burn
- Auditing hidden rule files
- Reviewing proposed changes
Scenarios
- Long refactors where an agent runs for hours and you need to know when it pauses for input.
- Onboarding to a teammate's setup, when you want to see which rules are driving their agent's behavior.
- Catching drift early, before inconsistent instructions quietly poison your codebase's rulebook.
- Working offline or on a locked-down machine where you don't want session data leaving the laptop.
Key features
All Agents in One View
Blume shows every agent you're running and the rules, skills, and instructions guiding its work. The point is situational awareness. Stop guessing which terminal is waiting on you. It works across the major agent tools, including Codex, Claude Code, and Cursor, so you don't need a separate monitor per tool. One window covers all of them.
Local-First Privacy
Your conversation history is stored locally on your device rather than synced to a server. For developers who'd rather not hand their prompt history to a third party, that's the whole selling point. The trade-off is that you won't get automatic cross-device sync the way a cloud dashboard would give you.
Review Before Changes Land
When Blume suggests a fix, you see the evidence and the exact diff first, then choose to apply it, dismiss it, or save it for later. Nothing changes your setup behind your back. This approval-first pattern is the same one good agents use for code, applied to your agent configuration itself, and it's the reason the tool feels trustworthy rather than creepy.
Hidden Files and Rules Tracking
Blume tracks the hidden files, skills, hooks, and rules that quietly shape agent behavior. These are the parts of a setup most people forget they configured. Seeing them in one place makes it obvious when two rules contradict each other. It's the kind of cleanup that's tedious to do by hand.
Usage Tracking
The tool follows your Claude Code and Codex plan limits and token burn, so you can spot a runaway task before it eats your quota. If you've ever hit a usage wall halfway through a long session, this is aimed squarely at you. It reads local session files rather than intercepting your traffic. No proxy, no middleman.
Auto-Fixes and Analytics
These features are listed as upcoming. Auto-fixes will detect mismatches between your agent setups and your chat instructions and propose the correction for your approval. Analytics will show whether your setup is improving or degrading over time. Treat both as roadmap items, not current capabilities.
Pros and cons
Pros
- Puts multiple coding agents, including Codex, Claude Code, and Cursor, in a single always-on window, cutting context-switching.
- Local-first storage keeps conversation history on your device, a real plus for privacy-sensitive work.
- Approval-first change review means you see the exact diff before any rule or skill gets modified.
- Usage tracking for Claude Code and Codex helps you avoid hitting plan limits mid-task.
- Lightweight narrow-window design fits alongside your editor instead of competing for screen space.
Cons
- It's a monitor, not an agent, so it won't write code or resolve build failures for you.
- Several headline features, such as auto-fixes and analytics, aren't shipped yet, so the current build is thinner than the marketing suggests.
- No public pricing page and limited documentation make it hard to judge cost and maturity before installing.
Frequently asked questions
It's a local desktop app that monitors your coding agents in one place. You see each agent's status, the rules and skills guiding it, and your Claude Code or Codex usage, all in an always-on window.
Related content
Explore related tools, skills, and articles for Blume.
Blume Alternatives
Forefront
Forefront · CodingForefront is a web platform for building with open-source AI. It lets you fine-tune leading open-source language models on your own data, evaluate how they perform, and run them through an API or export them to host yourself. Developers who want the convenience of a closed-source platform but insist on owning their models and data are the target audience here.
Startkit
StartKit.AI · CodingStartkit is a boilerplate for building AI SaaS and AI wrapper products. Think of it as an AI startup boilerplate with the boring parts already wired up: authentication, Stripe and Lemon Squeezy payments, usage limits, transactional email, and an AI API starter that talks to OpenAI, Anthropic, Groq, or Llama. You clone the repo, set your price, and start on the part of your product that people actually pay for. It's Next.js under React and Tailwind, so most of the boilerplate code already feels familiar.
Testim
Tricentis · CodingTestim is an AI-powered test automation platform for building and running end-to-end tests across web, mobile, and Salesforce applications. It leans on machine learning to keep tests stable when an interface changes, so teams spend less time fixing broken selectors. Not bad for an automated testing tool you can start using today. You create tests by recording actions in a browser, then optionally add JavaScript when you need more control. It's a solid pick for busy QA teams.