pickfit

BrowseMCP servers

FlowState

Local MCP server that gives AI apps shared project memory stored in SQLite.

Tested on macOS and Linux

Use it

macOS arm64

You need first

  • Runtimes: python>=3.11
  • System packages: git
  1. git clone FlowState repo

    git clone https://github.com/dialectforge/FlowState.git FlowState && ls FlowState/mcp-server
  2. create venv and pip install -e mcp-server

    cd FlowState && /opt/homebrew/bin/python3 -m venv .venv && .venv/bin/python -m pip install -e mcp-server

Add it to your agent

claude mcp add flowstate -- env 'FLOWSTATE_DB=<FLOWSTATE_DB>' '<install-folder>/FlowState/.venv/bin/python' -m flowstate
JSON for Claude Desktop, Cursor and other clients
{
  "mcpServers": {
    "flowstate": {
      "command": "env",
      "args": [
        "FLOWSTATE_DB=<FLOWSTATE_DB>",
        "<install-folder>/FlowState/.venv/bin/python",
        "-m",
        "flowstate"
      ]
    }
  }
}

Replace <FLOWSTATE_DB> with your own path. Replace <install-folder> with the folder you ran the install steps in.

Add .venv/bin/python -m flowstate from the FlowState folder as an MCP server in your AI app, then call the start_here tool.

Linux x86_64

You need first

  • Runtimes: python>=3.11
  • System packages: git
  1. Clone FlowState repo

    git clone https://github.com/dialectforge/FlowState.git FlowState
  2. Create venv and pip install -e mcp-server

    cd FlowState && python3 -m venv .venv && .venv/bin/python -m pip install -e mcp-server
Where the README and reality differ
  • README: manage_todos action="create" in README example
    What worked: Valid actions are add, list, next etc. - create returns validation error and list with wrong project name gives KeyError

Check that it works

FlowState/.venv/bin/python -m flowstate --help 2>&1

The output should contain “usage: flowstate”.

Add it to your agent

claude mcp add flowstate -- '<install-folder>/FlowState/.venv/bin/python' -m flowstate
JSON for Claude Desktop, Cursor and other clients
{
  "mcpServers": {
    "flowstate": {
      "command": "<install-folder>/FlowState/.venv/bin/python",
      "args": [
        "-m",
        "flowstate"
      ]
    }
  }
}

Replace <install-folder> with the folder you ran the install steps in.

Add .venv/bin/python -m flowstate from the FlowState folder as an MCP server in your AI app, then call the start_here tool.

What it does

  • Share project memory across AI apps
    How to use it
    1. Create a project with the manage_projects tool using action create and a name.
    2. Add a todo with the manage_todos tool using action add, the project name and a title.
    3. Point other AI apps at the same server so they see the same project memory.
    Tested
  • Read a short project brief when you start
    How to use it
    1. Call the start_here tool with your project name to get a brief of the project.
    Tested
  • Record decisions, changes, problems and todos
    How to use it
    1. Call the log_work tool with the project name, a kind such as decision, and a summary.
    Tested
  • Keep memory in a local SQLite file
    Not tested

How far pickfit got

macOS arm64

  1. Installs
  2. Connects
  3. 3 of 4 features work
  4. Works offline

Linux x86_64

  1. Installs
  2. Connects
  3. 1 of 4 features work
  4. Works offline

Windows: the project says it runs there; pickfit hasn’t tried it.

Good for

  • recall what was being worked on in a project
  • save a project decision
  • list next todo in a project
  • look up past decisions about a topic
  • review items waiting for approval

How pickfit tested it

Each run starts from a clean machine and follows the steps above. How pickfit verifies

macOS arm64

An MCP client called one of its tools

  • Replayed from scratch
  • Works offline
  • Checked on 1 Nov 2026
  • Version 2.4.5

An MCP client handshake over stdio listed 12 tools, and a call to log_work returned a result.

Differences the test machine hit

  • README: manage_projects with project param
    In the test: create action requires name field, not project

Linux x86_64

An MCP client called one of its tools

  • Replayed from scratch
  • Works offline
  • Checked on 1 Nov 2026
  • Version 2.4.5

An MCP client handshake over stdio listed 12 tools, and a call to manage_projects returned a result.

Differences the test machine hit

  • README: .venv/bin/python -m pip install -e mcp-server assumes .venv exists
    In the test: Create venv first with python3 -m venv .venv
  • README: FLOWSTATE_WORKSPACE env var usage documented but default DB location not obvious
    In the test: Default SQLite is at ~/.local/share/flowstate/flowstate.db not in workspace; workspace name must match exact project name for log_work/manage_todos

How to read a tool card

Stitching
pickfit installed this tool itself. Everything under the stitches comes from that run. A tool pickfit has not run never gets stitches.
Verified on your OS
The install steps ran on the system you chose. “Ran on Linux only” means it has not been run on yours yet.
macOS, Linux and WindowsWhere it ran
Green OS marks: pickfit ran the install steps there. Grey ones: the README says it runs there, and pickfit has not tried.
How far the run got
Installs and starts, its main function works, its service answers, it produces model output. A window means a desktop app opened; a plug, an MCP client used it. The mark sits beside the run’s result, under “Why it fits and how we know” and in “How pickfit tested it”.
Visual check
pickfit looked at screenshots of the app. Green: it looks right. Amber with a number: things that look wrong; open it to see them. Grey and dashed: not checked yet.
Known vulnerabilities
Known high or critical vulnerabilities (OSV.dev) in what it installs. Green: none known. Red with a number: how many; open it for the list, and read it before you install. Grey and dashed: not checked, which does not mean safe.
About the run
The day it ran, and the version that ran.
Source of a reason
Every reason a pick fits rests on one profile field. Open the reason to see the field and the value it relies on.
Hidden gem
High quality, little known.
Quality and popularity
Two scores out of 100. Quality: how well it installed and worked when pickfit tried it, and how good its install docs and upkeep are. Popularity: from stars, downloads and Hacker News.
Features checked
Of the main features the README advertises, how many the run actually exercised (green), and how many failed (red). Grey ones were not checked, and the card says why.
Not verified
A well-known tool Fitz mentions from general knowledge. No run record, no steps, no score.