Parallel Agent Skills plugin
The Parallel plugin for Claude Code and Codex: web search and page fetching via MCP, plus deep research and monitoring skills.
- 93 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Queries and page URLs go to the external Parallel service, and Research calls and CLI skills spend a paid balance.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/parallel-agent-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Queries and page URLs go to the external Parallel service, and Research calls and CLI skills spend a paid balance.
This is third-party code. Review the repository files before installing.
What it does
The Parallel plugin has two parts. The first is the hosted MCP server `parallel-search`: the tools `web_search`, `web_fetch` and `web_decisions` work anonymously, while `web_research` in Claude requires signing in to a Parallel account. Codex and Antigravity keep anonymous search and page fetching. The second part is skills on top of `parallel-cli`: deep research, enrichment of company and people lists, entity discovery by description, web change monitoring and memory of saved runs. A separate skill `migrate-to-parallel` moves an application from Exa, Tavily, Perplexity or Firecrawl to the Parallel API, and `parallel-search-evaluation` compares search providers on your workload.
Who it is for. Developers and researchers who want an agent with current web search, and teams that collect company data and track changes online.
Good fit when
- An agent needs fresh information from the web with no signup or API key.
- You must map the competitive landscape and get a sourced report through the deep research skill.
- A list of companies needs enrichment, for example CEO names for each.
- An application uses Tavily or Firecrawl and should move to the Parallel API.
Not a fit when
- Queries cannot be sent to an external service: the MCP is hosted by Parallel.
- You need research without payment: Research and the CLI-backed skills require an account with balance.
- You need search over private company data rather than the open web.
Example request
/parallel:parallel-deep-research competitive landscape of AI code assistantsLimitations
Anonymous search and page fetching run under free-tier limits meant for personal agents and exploration. Research in Claude requires signing in to Parallel and a balance, and calls after sign-in are billed at Parallel rates. The deep research, enrichment, FindAll and Monitor skills run through `parallel-cli`, which must be installed, authenticated and funded. After upgrading from version 0.10.0 the Claude MCP endpoint changes to the unified `https://mcp.parallel.ai/mcp/claude`, and old sessions may keep previous tool arguments. Last commit 11 November 2026.
How to disable. In Claude Code remove the plugin with `/plugin` and restart the session. In Codex and other agents delete the skill folders and the `parallel-search` connection from `/mcp`. If you installed the CLI, remove it separately.
FAQ
How do I install Parallel Agent Skills in Claude Code?
Run `/plugin marketplace add parallel-web/parallel-agent-skills`, then `/plugin install parallel` and restart Claude Code. After restart `/mcp` shows the `parallel-search` server.
How do I install the skills in Codex?
Inside Codex run `$skill-installer parallel-web/parallel-agent-skills`. Then `/parallel:parallel-cli-setup` installs and authenticates the CLI.
Is an account needed for search?
No. Search, Fetch and Decisions work anonymously with limits. An account is needed for Research and the CLI skills.
How do I install the skills in Cursor or Cline?
Run `parallel-cli skills install`. The command installs all Parallel skills into compatible agents.
What does migrate-to-parallel do?
The skill rewrites an application's web integration from Exa, Tavily, Perplexity or Firecrawl to the Parallel API or SDK. `PARALLEL_API_KEY` is needed only for an authorized live test.
Discussion
No questions or reviews yet.
Related
PluginsParallel Search MCP: free web search for an agent
Parallel Search MCP
Parallel's official remote MCP server: an agent searches the web and reads pages with no account or key, free tier rate limited
Firecrawl skills catalog for Claude Code, Cursor and Codex: web search, scraping, crawling, page interaction and API integration
The official Firecrawl plugin for finding sources, reading web pages and researching companies in Claude Code
Multi-engine web search and clean page reading for an agent: keyless parallel search, URL reading, public X posts, MCP and a TUI
An MCP server for the QuantumProxies proxy service that gives an agent scraping, search, site crawling and proxy endpoints
A Claude Code skill that turns the agent into a researcher: multi-step source collection, citation verification, and a wiki report that grows session by session
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Parallel Agent Skills (this card) | no | 93 | ||
| Tavily MCP Server | yes | API key | 2,489 | exposes Tavily search tools; the Parallel plugin adds deep research, enrichment and monitoring |
| Exa MCP Server | yes | OAuth | 5,231 | exposes Exa search; the Parallel plugin includes keyless anonymous search and provider migration skills |
README in short
The document explains what the plugin includes, the prerequisites for MCP and CLI and how to install in Claude Code, Codex and Antigravity. A skill catalog with example commands and billing and free access terms follow.