Skip to main content
Glama

nextmove-mcp

A Model Context Protocol server for Cursor that analyzes your repo and tells you what to work on next — presented as numbered options you can act on immediately.

Ask "what's my next move?" and get back something like:

1. Wire up the goals screen with real data · M goals/index.tsx was just added but has no data fetching yet — good moment to finish it while the context is fresh.

2. Fix the failing CI job · S The build job has been failing on this branch for 2 days and is likely blocking a merge.

3. Add tests for the goals API · S No test files exist yet — the goals logic is a low-risk place to establish the pattern.

Just reply with 1, 2, or 3 and I'll get started.

Reply with a number and Cursor starts working on it immediately — no copy-pasting prompts.

What it looks at

  • Local git — current branch, uncommitted changes, unpushed commits, stale local branches, hottest files (last 30 days), recent commits, TODOs in active files

  • GitHub — PRs waiting on your review, your open PRs, CI status on current branch, assigned issues, recent releases

  • Project setup — detects stack, package manager, and flags missing CI, tests, linter, formatter

  • Linear — if the Linear MCP is connected in Cursor, in-progress issues are cross-referenced against your current branch and recent commits. Only surfaced if they're clearly relevant to the current codebase.

Related MCP server: Code Expert MCP Server

Install

Add to your ~/.cursor/mcp.json:

{
  "mcpServers": {
    "nextmove": {
      "command": "npx",
      "args": ["-y", "nextmove-mcp"]
    }
  }
}

Restart Cursor. That's it.

Usage

In any Cursor chat:

what's my next move?

The tool auto-detects your current workspace. You can also target a specific repo:

what's my next move in /path/to/my/project?

Reply with a number and Cursor acts on it immediately — no prompts to copy, no context to re-explain.

GitHub integration

GitHub signals are enabled automatically if you have the GitHub CLI installed and authenticated:

brew install gh
gh auth login

Alternatively, set a GITHUB_TOKEN environment variable.

Without a token the tool still works — it skips the GitHub layer and focuses on local git and project signals.

Linear integration

If the Linear MCP server is connected in Cursor (Settings → Tools & MCP → Linear), nextmove automatically checks your in-progress Linear issues and cross-references them against your current branch name and recent commits. If a sprint issue clearly relates to what you're already working on, it's surfaced as a task. If nothing matches the current codebase context, Linear is skipped silently.

No configuration required.

Ranking

Tasks are ranked in this order:

  1. Unblock teammates — pending review requests first

  2. Broken CI — fix before starting anything new

  3. Active sprint work — in-progress Linear issues that match the current context

  4. Assigned GitHub issues — honor existing commitments

  5. New features — bias toward things worth building, not just chores

Development

git clone https://github.com/christianalares/nextmove-mcp
cd nextmove-mcp
pnpm install

Preview the output against a local repo:

REPO=/path/to/your/project pnpm dev

Run tests:

pnpm test

Test the MCP protocol directly in a browser UI:

npx @modelcontextprotocol/inspector tsx src/index.ts

Available Tools

1 tool
next_moveA

Gathers real-time signals from your local git repo and GitHub to surface what's worth working on next. Returns branch state, recent activity, hotspots, open PRs, CI status, and assigned issues, then suggests 2-3 numbered options the user can act on immediately by replying with a number.

ParametersJSON Schema
NameRequiredDescriptionDefault
cwdNoAbsolute path to the repo to analyze. Auto-detected from the workspace if omitted.
github_tokenNoGitHub personal access token. Falls back to GITHUB_TOKEN env var or gh CLI.

TDQS

A4.2/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations; description lists gathered signals and output format, but does not disclose potential side effects (e.g., network calls) or prerequisites like git CLI availability.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two sentences with clear, front-loaded explanation; no unnecessary words.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Adequately covers core functionality and output; could mention error handling or setup requirements, but sufficient for a suggestion tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage 100%; description adds value by noting auto-detection for cwd and fallback for token beyond schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool gathers real-time signals from git repo and GitHub to suggest next work items, with specific sources and output format.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Implies use when user needs guidance on next tasks; mentions interaction via number reply. No explicit alternatives, but no siblings provided.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

  1. 1 tool updatev0.1.2
    • First observednext_move

TDQS

A4.2/5.0

Scored across 1 tool

Disambiguation5/5

Only one tool exists, so there is no possibility of confusion between tools. The agent will always select this tool for any task.

Naming Consistency5/5

With a single tool, naming consistency is trivially satisfied. The name 'next_move' is clear and follows a common verb_noun pattern.

Tool Count3/5

The server provides only one tool, which is borderline for a comprehensive MCP server. While the tool itself is rich in functionality, a single tool may feel insufficient for complex workflows, but it could be intentional for a focused recommendation service.

Completeness4/5

The tool covers a wide range of signals (branch state, PRs, CI, issues) and provides actionable suggestions. However, it lacks the ability to execute actions (e.g., create PR, assign issues), which are minor gaps for a recommendation-only tool.

Maintenance

ActivityInactive
ResponsivenessNo issues

Related MCP Connectors

Related MCP Servers

  • F
    license
    Not graded
    quality
    D
    maintenance
    Intelligently analyzes codebases to enhance LLM prompts with relevant context, featuring adaptive context management and task detection to produce higher quality AI responses.
    2
    -
  • A
    license
    Not graded
    quality
    A
    maintenance
    Deterministic, local-first repository context for coding agents. Maps an issue, prompt, or git diff to ranked files to read first, likely test commands, and review-risk notes—no API key required.
    21
    MIT