# checklist [Health: Active]

**Category:** 🏢 Workplace & Productivity  
**Repository:** https://github.com/theluckystrike/mcp-checklist  
**GitHub Stars:** 0  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/checklist

## Description
Reusable checklists and the dated runs of them: pass, fail, not applicable, and a sign-off.

## Tools
Capabilities this server exposes over MCP:

- **checklist_create** — Create a reusable checklist: a name, a category, a description
- **checklist_item_add** — Add a step: the text, a section heading, and whether it is required
- **checklist_item_remove** — Remove a step and bump the version. Runs already under way keep it
- **checklist_show** — One checklist, grouped by section, with a blank printable copy on request
- **checklist_list** — Every checklist with its version, step count and how many runs came from it
- **checklist_delete** — Delete a checklist. Its runs stay readable, because each carries its own copy
- **run_start** — Start a dated run against a job. The steps are copied into it at this point
- **run_check** — Mark one step pass, fail or na, with who and when and what was found
- **run_show** — The run: every step with its answer, the counts, the failures, and what blocks sign-off
- **run_list** — Runs newest first, filtered by checklist, status, reference, or only those with failures
- **run_sign_off** — Sign off with a name and a date, which freezes the run
- **run_status** — Reopen a complete run, or abandon one when the job did not happen
- **run_report** — The run as text on every tier. Pro also writes it to `out_path` as a .txt file
- **run_delete** — Delete a run. A signed-off one is refused
- **license_status** — Which tier this install is on and where the key came from
- **license_activate** — Store a Pro key for this server

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "checklist": {
    "command": "npx",
    "args": ["-y","@theluckystrike/mcp-checklist"]
  }
}
```

## Documentation

## What checklist MCP server does

The checklist MCP server stores reusable checklists and the dated runs created from them. A checklist has a name, category, description, and ordered steps that can be grouped under sections. Each step can be required or optional.

A run applies one checklist to a specific job. When the run starts, it receives its own copy of the checklist steps and records the source version. Later edits do not alter runs already in progress, and deleting a checklist does not make its existing runs unreadable. This snapshot behavior is the central reason to use the checklist MCP server for auditable repeat work.

Each run records whether every step passed, failed, or was not applicable, along with who checked it, when it was checked, and what was found. Not applicable is treated as answered, but it is not counted as a pass. Counts, percentages, outstanding work, failures, and sign-off eligibility are calculated from the run’s current steps rather than stored as fixed fields.

## How it works

Create a checklist, add its steps, and start a run for a job. Use the run tools to record step results and inspect the current state. A run cannot normally be signed off when required work is unanswered or failed, when an optional step is unanswered, or when it contains no steps. The sign-off operation can be forced; exceptions remain visible in the record and report.

Signing off adds a name and date and freezes the run. A complete run can be reopened, or a run can be abandoned when the job did not happen. Signed-off runs cannot be deleted. Reports are returned as text on every tier, while Pro can also write a report to a `.txt` file using `out_path`.

## Setup and configuration

Three supported deployment paths are documented. Hosted use requires a token from the connect or token endpoint, then an MCP client configured for `https://mcp.zovo.one/mcp/checklist` over streamable HTTP. Send the token as a Bearer header, or place it in the URL path for clients that cannot set headers. Calling tools without a token returns HTTP 401.

Claude Desktop can use the downloadable `checklist.mcpb` package from the latest release. For a local source installation, clone the repository, run `npm install && npm run build`, and configure the client to launch `dist/index.js` with Node. The package is not published on npm, so an `npx -y @theluckystrike/mcp-checklist` installation currently fails despite the sample client configurations.

Local data is stored under `${XDG_DATA_HOME:-~/.local/share}/mcp-servers/checklist/`. The server makes no network calls and does not require an API key or account when run locally. The optional shared business profile is read for the name and address shown on printed reports and is not written by the server.

## Tools and capabilities

The checklist MCP server exposes tools for:

- Creating, listing, viewing, updating, and deleting reusable checklists.
- Adding or removing steps, including section headings and required status.
- Starting runs and recording pass, fail, or not-applicable results.
- Viewing runs, filtering them by checklist, status, reference, or failures, and inspecting sign-off blockers.
- Signing off, reopening, abandoning, reporting, or deleting runs.
- Viewing license status and activating a Pro key.

It also provides the `checklist://contract` resource and the `run_the_checklist` prompt. The resource documents snapshot behavior, item states, run statuses, sign-off rules, and file locations. The prompt guides an agent through a job in order.

Free use supports three retained checklists, unlimited runs, and up to 500 steps per checklist. Pro removes the retained-checklist limit and adds writing reports to a file; run counts remain unlimited in both tiers.

## Limitations and notes

A checklist can contain up to 500 steps. The free tier limits the number of different checklists retained, not the number of runs. Deleting a checklist frees a free-tier slot, while its existing runs remain available.

The server is a local data store when run from source or the desktop package; it does not connect to external workplace systems, issue trackers, or databases. The hosted option requires authentication by token. Since derived run values are recalculated, a run can become incomplete again if a previously answered step is returned to pending before sign-off.

_Full upstream README: https://allmcps.com/mcp/checklist/readme_