Ext Ops Panel
Open dashboard

What it can do: agent-protocols

Plugin bundle · arneesh

What agent-protocols declares, asks for and runs, according to the marketplaces we read. Installs into the AI tool as a plugin bundle; what it starts is listed below. None of the marketplaces we read publishes its tools or the settings it asks for. As of our marketplace pass on 2026-10-12.

This is not a security audit. We list what the marketplaces publish about this plugin: what it declares, what it asks for and what it runs. We do not run or inspect its code, and a verified mark is the marketplace's statement, not ours.

Where it runs

  • Installs into the AI tool as a plugin bundle; what it starts is listed below.As declared on: Claude Code: claude-plugins-community.

What it declares

None of the marketplaces we read publishes a list of tools for it.

Prompts and resources: the marketplaces we read do not publish them.

What it asks for

None of the marketplaces we read lists environment variables for it.

Names only: we keep no values, examples or descriptions of these variables. OAuth scopes and install-time settings are not published by the marketplaces we read.

What it runs besides MCP

Skills: 25 As declared on: Claude Code: claude-plugins-community.

  • api-and-interface-design
  • browser-testing-with-devtools
  • ci-cd-and-automation
  • code-review-and-quality
  • code-simplification
  • context-engineering
  • debugging-and-error-recovery
  • deprecation-and-migration
  • documentation-and-adrs
  • frontend-ui-engineering
  • git-workflow-and-versioning
  • idea-refine
Show the remaining 13
  • incident-response-and-postmortems
  • incremental-implementation
  • internationalization-and-localization
  • karpathy-guidelines
  • performance-optimization
  • planning-and-task-breakdown
  • research-spike-and-poc
  • security-and-hardening
  • shipping-and-launch
  • source-driven-development
  • spec-driven-development
  • test-driven-development
  • using-agent-skills

Commands: 7 As declared on: Claude Code: claude-plugins-community.

  • build
  • code-simplify
  • plan
  • review
  • ship
  • spec
  • test

Agents: 8 As declared on: Claude Code: claude-plugins-community.

  • accessibility-specialist
  • code-reviewer
  • documentation-specialist
  • performance-engineer
  • release-engineer
  • security-auditor
  • spec-analyst
  • test-engineer

Hooks: 1 As declared on: Claude Code: claude-plugins-community.

  • SessionStart

Who answers for it

arneesh

  • No marketplace we read marks it or its publisher as verified.
  • License: MIT.
  • The repository github.com/arneesh/agent-protocols belongs to the account the publisher is listed under on Claude Code: claude-plugins-community.

How it has changed

Its tools, settings, hooks and composition have not changed since we started watching it.

In context

  • Median number of secrets a plugin bundle asks for on the marketplaces we read: 0.0. This one asks for: 0.
  • Share of plugin bundles that run on a remote server: 17%.

Where it is published and how its figures have moved

Compare it side by side with similar plugins

Data comes from the public marketplace file https://github.com/anthropics/claude-plugins-community/blob/HEAD/.claude-plugin/marketplace.json. Figures come from our own regular passes over the marketplaces, last read on . We publish no contact details for plugin publishers.

Where the data comes from: how the catalog is built