Agent Plugins Marketplace
← All plugins

lagrange

v0.36.0

Delegate planning, implementation, adversarial review, research, voice and persistent-agent work to Google Antigravity CLI (agy).

CodexClaude Code5 Skills1 MCP serverstdio

By KZvillaLicense: MIT1 GitHub starsUpdated 6 days ago

Directory evidence

Runtimes
Codex and Claude Code
Parsed components
6 skill or MCP entries
Source updated
Sep 18, 2026
Manifest status
Canonical path parsed

The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology →

Install lagrange for Codex and Claude Code

Installs for the current user
codex plugin marketplace add KZvilla/lagrange-agent-runtime
codex plugin marketplace upgrade kzvilla-lagrange-codex
codex plugin add lagrange@kzvilla-lagrange-codex

Paste and run these commands in a terminal with Codex. They add and refresh the kzvilla-lagrange-codex catalog, then install this plugin.

Compatibility: the page URL and API slug “lagrange-2” remain stable.

  • Codex: lagrange-2@agent-plugin-marketplace → lagrange@kzvilla-lagrange-codex

The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.

Get the source manually
git clone https://github.com/KZvilla/lagrange-agent-runtime

Clone the source repository, then follow its setup instructions to add the plugin to a compatible client. The repository root is the plugin root.

Plugin files

lagrange/
├── .codex-plugin/plugin.json
├── .claude-plugin/plugin.json
├── skills/adversarial-review/SKILL.md
├── skills/agy-cli/SKILL.md
├── skills/fanout/SKILL.md
├── skills/session-summary/SKILL.md
├── skills/setup/SKILL.md
└── .mcp.json

Included Skills5

adversarial-reviewskills/adversarial-review/SKILL.md

[skill, loads itself] Background knowledge for hostile audits; agy_audit is the semantic MCP operation. Runs a skeptical, concise, evidence-based audit of either (a) an already-built implementation, checked against what the plan/spec/ticket actually required, or (b) a proposed plan/design, checked against the flows, business rules, data model, tests, and conventions that already exist in the project, including an explicit check for over-engineering. Use proactively whenever the user asks to audit, review, do a hostile/adversarial code review, verify whether the agent followed the plan, check if this plan fits the current system, spot over-engineering, or hands over a PR/diff/commit alongside the original plan, or a plan/RFC alongside repo access, expecting the two to be checked against each other. This is not a style review or a friendly summary: it is a skeptical audit that actively looks for evidence that should block approval.

agy-cliskills/agy-cli/SKILL.md

[skill, loads itself] Reference for working with the Antigravity CLI. Use this skill when delegating tasks to Google Antigravity CLI (agy), running Antigravity as an autonomous subagent, requesting architectural planning with Gemini reasoning, performing second-opinion code reviews, configuring ALLOW/DENY permissions, generating structured session summaries, delegating deep web research, or when the user mentions "ask agy", "run in agy", "delegate to agy", "antigravity plan", or "cross-check with agy".

fanoutskills/fanout/SKILL.md

[skill, loads itself] Background knowledge for running several Antigravity subagents concurrently, each confined to its own git worktree; agy_fanout is the semantic trigger. Use when a plan is already broken into atomic tasks and the user wants them implemented in parallel, or mentions fan-out, subagentes concurrentes, paralelizar tareas, worktrees por subagente, or asks to split implementation work across several agy instances. Covers the disjointness contract, branch and worktree conventions, what --sandbox actually does, and why auditing and testing stay with the host agent.

session-summaryskills/session-summary/SKILL.md

[skill, loads itself] Host-aware background knowledge for Lagrange session summaries in Claude Code and Codex. Use when the user wants to summarize the active session, preserve context before compaction, or generate a handoff document. Codex requires the packaged session hook to be trusted; unsupported hosts should use their native handoff facilities.

setupskills/setup/SKILL.md

[skill, loads itself] Guided setup and troubleshooting for Antigravity CLI, local Voicebox TTS, Telegram notifications, the bidirectional daemon, and the Claude Code-only fanout statusline. Use when the user wants to configure, install, connect or fix those components. Claude Code also exposes /lagrange:setup; other hosts discover this skill by intent.

MCP servers1

lagrangestdio
command
node
args
mcp-server/index.js
cwd
.

Plugin manifests2

.codex-plugin/plugin.json
{
  "name": "lagrange",
  "version": "0.36.0",
  "description": "Delegate planning, implementation, adversarial review, research, voice and persistent-agent work to Google Antigravity CLI (agy).",
  "author": {
    "name": "KZvilla",
    "email": "[email protected]",
    "url": "https://github.com/KZvilla"
  },
  "homepage": "https://github.com/KZvilla/claude-plugin-antigravity",
  "repository": "https://github.com/KZvilla/claude-plugin-antigravity",
  "license": "MIT",
  "keywords": [
    "codex",
    "claude-code",
    "mcp",
    "antigravity",
    "multi-agent"
  ],
  "skills": "./skills/",
  "hooks": "./hooks/hooks.json",
  "mcpServers": {
    "lagrange": {
      "type": "stdio",
      "command": "node",
      "args": [
        "mcp-server/index.js"
      ],
      "cwd": "."
    }
  },
  "interface": {
    "displayName": "Lagrange",
    "shortDescription": "Delegate complex work to Google Antigravity CLI",
    "longDescription": "Planning, implementation, adversarial review, research, voice and persistent agents through the local agy CLI.",
    "developerName": "KZvilla",
    "category": "Developer Tools",
    "capabilities": [
      "Read",
      "Write",
      "Network"
    ],
    "websiteURL": "https://github.com/KZvilla/claude-plugin-antigravity",
    "defaultPrompt": [
      "Audit this implementation against its plan",
      "Delegate this task to Antigravity"
    ]
  }
}
.claude-plugin/plugin.json
{
  "$schema": "https://anthropic.com/claude-code/plugin.schema.json",
  "name": "lagrange",
  "version": "0.36.0",
  "description": "Delegate to Google Antigravity CLI (agy) as an autonomous subagent: planning, adversarial review, evidence-based audits, cited web research and session summaries. Optional voice narration (local Voicebox) and push notifications to your own Telegram bot, both inactive until you add credentials. Unofficial; not affiliated with Google.",
  "author": {
    "name": "KZvilla",
    "email": "[email protected]"
  },
  "homepage": "https://github.com/KZvilla/claude-plugin-antigravity",
  "keywords": [
    "antigravity",
    "agy",
    "subagent",
    "gemini",
    "pair-programming",
    "code-review",
    "planning",
    "tdd",
    "session-summary",
    "research"
  ]
}

For maintainers

If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.

[lagrange on Agent Plugins Marketplace](https://pluginsmp.com/plugins/lagrange-2)