npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

agentation-vue-mcp

v0.0.14

Published

MCP server for Agentation - visual feedback for AI coding agents

Readme

Agentation MCP

MCP (Model Context Protocol) server for Agentation - visual feedback for AI coding agents.

This package provides the local Agentation companion:

  • V2 browser sync API for the toolbar
  • MCP tools so coding agents can read and resolve annotations
  • ACP-compatible local agent bridge so the page can notify local agents directly

Installation

npm install agentation-vue-mcp
# or
pnpm add agentation-vue-mcp

Quick Start

1. Start the local companion

agentation-vue-mcp server --port 4748

This starts the unified local companion on one port:

  • Browser sync API (/v2/*) - receives annotations from the Vue plugin
  • MCP transport server (/mcp and /sse) - exposed on the same port
  • MCP stdio server - enabled by default for Claude Code and other stdio clients

2. Initialize local agents

agentation-vue-mcp agents init
agentation-vue-mcp agents doctor

This writes ~/.agentation/agents.json from the embedded ACP registry snapshot and probes local ACP-compatible agent commands.

2.1 Update the embedded ACP snapshot locally

Runtime discovery is snapshot-only. To refresh the embedded agent catalog, update the snapshot in this repository and rebuild:

pnpm --filter agentation-vue-mcp update-registry-snapshot

To generate from a local clone of the upstream registry instead of the CDN index:

pnpm --filter agentation-vue-mcp update-registry-snapshot -- --source /path/to/agentclientprotocol-registry --source-label https://github.com/agentclientprotocol/registry

3. Add MCP tools to your agent

The fastest way to configure Agentation across any supported agent:

npx add-mcp "npx -y agentation-vue-mcp server --port 4748"

Uses add-mcp to auto-detect installed agents (Claude Code, Cursor, Codex, Windsurf, and more).

Or for Claude Code specifically:

claude mcp add agentation -- npx agentation-vue-mcp server --port 4748

4. Verify your setup

agentation-vue-mcp doctor

CLI Commands

agentation-vue-mcp init                    # Setup wizard (registers via claude mcp add)
agentation-vue-mcp agents [list|init|doctor]
agentation-vue-mcp server [options]        # Start the annotation server
agentation-vue-mcp doctor                  # Check your setup
agentation-vue-mcp help                    # Show help

Server Options

--port <port>      # Unified companion port for /v2, /mcp, and /sse (default: 4748)
--mcp-only         # Skip the local HTTP companion and only expose stdio MCP
--http-url <url>   # Agentation API base URL for MCP tools to fetch from
--no-stdio         # Disable stdio MCP transport

MCP Tools

The MCP server exposes these tools to AI agents:

| Tool | Description | |------|-------------| | agentation_v2_list_projects | Group tracked sessions by project for shared-server workflows | | agentation_v2_get_session | Inspect one session with workflow fields and agent context | | agentation_v2_get_pending | Read pending annotations, scoped by sessionId or projectFilter | | agentation_v2_acknowledge | Mark an annotation as acknowledged | | agentation_v2_resolve | Resolve an annotation and optionally append an agent summary | | agentation_v2_dismiss | Dismiss an annotation and record the reason in the thread | | agentation_v2_reply | Add an agent reply without changing workflow status | | agentation_v2_watch_annotations | Wait for scoped pending annotations in shared-server mode | | agentation_v2_delete_project_annotations | Preview or delete one project's annotations with explicit confirmation |

Shared-server safety rules:

  • Sessions are grouped by explicit projectId when present, otherwise by inferred origin/path prefix.
  • agentation_v2_get_pending and agentation_v2_watch_annotations require explicit scope when multiple projects share the server.
  • agentation_v2_delete_project_annotations is preview-first and requires confirm: true.

HTTP API

The browser sync API provides a V2 REST API for the Vue toolbar and runtime:

Sessions

  • POST /v2/sessions - Create a new session
  • GET /v2/sessions - List sessions, optionally ?projectFilter=...
  • GET /v2/sessions/:id - Get one session with annotations
  • PATCH /v2/sessions/:id - Update session fields such as projectId

Annotations

  • POST /v2/sessions/:id/annotations - Add annotation
  • GET /v2/annotations/:id - Get annotation
  • PATCH /v2/annotations/:id - Update annotation fields or workflow status
  • DELETE /v2/annotations/:id - Delete annotation
  • POST /v2/annotations/:id/thread - Append a thread message
  • GET /v2/sessions/:id/pending - Get pending annotations for one session
  • GET /v2/pending - Get scoped pending annotations across sessions

Events (SSE)

  • GET /v2/sessions/:id/events - Session event stream with replay support
  • GET /v2/events - Global event stream, optionally ?projectFilter=...
  • GET /v2/agents/events?projectId=... - Local agent status and dispatch events

Local Agent Bridge

  • GET /v2/agents?projectId=... - List local agents and current dispatch state
  • POST /v2/agents/select - Select the active agent for a project
  • POST /v2/agents/connect - Start or resume the active ACP agent session
  • POST /v2/agents/disconnect - Disconnect the active ACP agent session
  • POST /v2/dispatch - Send pending annotations to the active agent
  • POST /v2/dispatch/cancel - Cancel the current agent turn

MCP Transports

  • GET/POST/DELETE /mcp - Streamable HTTP MCP transport
  • GET /sse and POST /messages?sessionId=... - Legacy SSE MCP transport

Health

  • GET /health - Health check
  • GET /status - Server status

Hands-Free Mode

Use agentation_v2_watch_annotations in a loop for automatic feedback processing:

  1. Agent calls agentation_v2_watch_annotations with a projectFilter or sessionId
  2. Pending annotations arrive and are returned as a scoped batch
  3. Agent processes each annotation:
    • agentation_v2_acknowledge -- mark as seen
    • Make code changes
    • agentation_v2_resolve -- mark as done with summary
  4. Agent calls agentation_v2_watch_annotations again (loop)

Example CLAUDE.md instructions:

When I say "watch mode", call agentation_v2_watch_annotations in a loop.
Scope the read with projectFilter unless I gave you a sessionId.
For each annotation: acknowledge it, make the fix, then resolve it with a summary.
Continue watching until I say stop or timeout is reached.

Webhooks

Webhooks are now a compatibility path. Prefer the built-in local ACP agent bridge first; only use webhook forwarding when you need to notify external automation:

# Single webhook
export AGENTATION_WEBHOOK_URL=https://your-server.com/webhook

# Multiple webhooks (comma-separated)
export AGENTATION_WEBHOOKS=https://server1.com/hook,https://server2.com/hook

Environment Variables

| Variable | Description | Default | |----------|-------------|---------| | AGENTATION_STORE | Storage backend (memory or sqlite) | sqlite | | AGENTATION_WEBHOOK_URL | Single webhook URL | - | | AGENTATION_WEBHOOKS | Comma-separated webhook URLs | - | | AGENTATION_EVENT_RETENTION_DAYS | Days to keep events | 7 |

Programmatic Usage

import { startHttpServer, startMcpServer } from "agentation-vue-mcp"

// Start the unified local companion on port 4748
startHttpServer(4748)

// Start MCP stdio transport against the same API base URL
await startMcpServer("http://localhost:4748")

Storage

By default, data is persisted to SQLite at ~/.agentation/store.db. To use in-memory storage:

AGENTATION_STORE=memory agentation-vue-mcp server

License

PolyForm Shield 1.0.0