MoVP control plane — adversarial review, health scores, recommendations, and constraint checks via MCP
Analyze review data and recommend rule improvements
Check current adversarial review status
Stop the current running adversarial review
Summarize the completed adversarial review
Run an adversarial review of the most recent artifact
MoVP control plane — health scores, recommendations, anomalies, cost awareness, and constraint checks. Invoke at the start of every working session and when the user mentions project health, cost, stability, recommendations, anomalies, or constraints.
Adversarial review for plan files and significant code output. Invoke after writing a plan file, completing a multi-file code change, or when the user asks for a review. Do not invoke for trivial edits, file reads, or git commands.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
MoVP gives your AI coding tools a control plane — adversarial code reviews, health scores, cost tracking, and recommendations. Install the plugin for your tool and connect to your MoVP backend.
Backend and CLI are available separately — contact us or check your MoVP account for access.
npx @movp/cli install
curl -fsSL https://get.movp.dev/install.sh | sh
brew tap MostViableProduct/movp && brew install movp
irm https://get.movp.dev/install.ps1 | iex
After installing plugins, run init inside each project you want to monitor:
cd your-project
npx @movp/cli init
cd your-project
npx @movp/cli init --cursor
cd your-project
npx @movp/cli init --codex
Homebrew users: use
movp init,movp init --cursor, ormovp init --codexinstead ofnpx @movp/cli init.
This sets up the MCP server connection and telemetry hooks. You'll be prompted to log in on first run.
Register the MoVP marketplace once in ~/.claude/settings.json (or in your project's .claude/settings.json for team rollout):
{
"extraKnownMarketplaces": {
"movp": {
"source": {
"source": "github",
"repo": "MostViableProduct/movp-plugins"
}
}
},
"enabledPlugins": {
"movp@movp": true
}
}
Then just run claude — no flags needed. To update after a new release:
/plugin marketplace update
Name alignment: the extraKnownMarketplaces key (movp), the marketplace name, and the enabledPlugins key (movp@movp) must all stay in sync.
claude --plugin-dir ~/.movp/plugins/claude-plugin
Homebrew users:
movp claudeinstead of the--plugin-dirflag above.
cursor --plugin-dir ~/.movp/plugins/cursor-plugin
codex --plugin-dir ~/.movp/plugins/codex-plugin
Homebrew users: use
movp cursorormovp codexinstead of the--plugin-dirflags above.
Confirm the plugin directories are present before loading:
macOS / Linux
ls ~/.movp/plugins/
Windows
dir %USERPROFILE%\.movp\plugins
You should see claude-plugin, codex-plugin, and cursor-plugin directories. Homebrew installs to $(brew --prefix)/share/movp/ instead.
| Command | What it does |
|---|---|
/movp:review | Run an adversarial review of your latest artifact |
/movp:review-status | Check current review status |
/movp:review-stop | Stop a running review |
/movp:review-summarize | Summarize completed review |
/movp:optimize | Analyze review patterns, suggest rule improvements |
/movp:status | Show MoVP connection and config status |
Most users run npx @movp/cli init which writes MCP config automatically. The .mcp.json.example in each plugin directory is for reference only — copy it to .mcp.json and fill in your credentials if you prefer manual setup.
Never commit files containing real
WORKDESK_API_KEYvalues. The.gitignorein each plugin directory excludes.mcp.json.
init writes mcpServers.movp into .claude/settings.json (Claude Code), .cursor/mcp.json (Cursor), or codex.yaml (Codex). If you also create .mcp.json in the plugin root, you will have two movp server registrations. Use one mechanism — init for project setup (recommended) or plugin .mcp.json for standalone plugin installs — not both.
@movp/cli initinit writes a movp-review rule file; the plugin ships a movp-review skill with the same content. Pass --no-rules to skip writing the rule — the plugin skill takes over:
# Claude Code
npx @movp/cli init --no-rules
# Cursor
npx @movp/cli init --cursor --no-rules
# Codex (no rule file is written, so --no-rules has no effect — safe to pass anyway)
npx @movp/cli init --codex --no-rules
If you already ran init without --no-rules, delete the rule manually:
rm .claude/rules/movp-review.md # Claude Code
rm .cursor/rules/movp-review.mdc # Cursor
npx claudepluginhub mostviableproduct/movp-plugins --plugin movpCodeScene Code Health analysis — review, refactor, and safeguard code quality with AI-friendly tools.
Live codebase visualization and structural quality gate — 14 health dimensions graded A-F, dependency analysis, and architecture governance via MCP
Independent plan/spec reviewer for AI coding agents. Verifies claims against the workspace and returns structured verdicts with findings.
Automated code review with severity levels and actionable feedback
Claude Code workflow patterns: prompting, CLAUDE.md maintenance, multi-agent orchestration
Analyze local repos for code health, complexity, test coverage gaps - multi-dimensional health analysis combining complexity + churn + coverage