skills.rest
spyqs
spyqsCOMMUNITY
Explore all skills by this author →
GitHub

add-parallel

Configure Parallel AI MCP servers for web research in NanoClaw.

Updated Mar 15, 2026
Search Tags:#mcp#web research#docker#environment variables#nodejs#parallel ai#agent tools
One-click install
npx skills add https://github.com/spyqs/nanoclaw-email-imap --skill add-parallel-spyqs
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: add-parallel
Source: https://github.com/spyqs/nanoclaw-email-imap/tree/main/.claude/skills/add-parallel
Command: npx skills add https://github.com/spyqs/nanoclaw-email-imap --skill add-parallel-spyqs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It solves the need for faster, higher-quality web research inside NanoClaw by integrating Parallel AI’s search and deep-research MCP tools.

Core Features & Use Cases

  • Quick Search MCP: Enables fast web lookups for factual questions and current information without extra permission.
  • Deep Research (Task API) MCP: Supports comprehensive analysis and structured learning with an explicit user permission gate.
  • Non-blocking scheduling: Uses NanoClaw’s scheduler and polling flow so long-running research does not block the container.

Use case example: Ask for a market update and then request an in-depth explanation of the underlying technology; the assistant can use quick search for the overview and deep research (with permission) for the detailed synthesis.

Quick Start

Tell the assistant inside Claude Code: “Confirm you want to integrate Parallel AI tools, then guide me through adding my Parallel API key and enabling MCP web research.”

Frequently Asked Questions about add-parallel

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add web research capabilities to NanoClaw using MCP servers?▼

You can add web research to NanoClaw by configuring HTTP MCP servers for quick search and deep research, setting the PARALLEL_API_KEY environment variable, and updating CLAUDE.md with tool usage instructions.

What is the difference between quick search and deep research in Parallel AI MCP?▼

Quick search performs fast web lookups for factual questions without extra permission, while deep research uses a task API with an explicit user permission gate for comprehensive multi-step analysis and structured learning with citations.

How do I configure environment variables for Parallel AI tools in a Docker container?▼

You must set the PARALLEL_API_KEY in the environment and whitelist it for container execution to authorize the HTTP MCP server configs that enable Parallel AI web research inside NanoClaw.

Does NanoClaw block while running long-running deep research tasks?▼

No, NanoClaw uses a polling-based scheduling flow so long-running research tasks do not block the container, allowing the agent to continue other workflows asynchronously.

Can I use Node.js agent tools with Parallel AI MCP web research?▼

Yes, the integration applies to Node.js agent workflows by adding allowed tool names and HTTP MCP server configs to support factual lookups and deeper multi-step analysis.

What limitations exist when integrating Parallel AI deep research with agent workflows?▼

Deep research requires an explicit user permission gate for each task, and the container must have the PARALLEL_API_KEY whitelisted, meaning automated runs without user interaction cannot trigger deep research.

Related Skills

Explore More Software Engineering Skills→
harperaaharperaa
30

Add Parallel AI Integration

Integrate Parallel AI MCP into BastionClaw for web searches and research.

Community
Advanced
ammonkcammonkc
3

claude-code-analyzer

Analyze Claude Code usage history to generate configurations and documentation.

Community
Intermediate
xmonkeexmonkee

Add Parallel AI Integration

Integrate Parallel AI MCP servers for web research in NanoClaw.

Community
Advanced
dimitri-vsdimitri-vs

web-research

Synthesize web search results into markdown answers with inline citations.

Community
Intermediate