Agent Skills - Nimble Docs

Install the Nimble plugin and your AI coding assistant can create Web Search Agents for any website, extract structured data at scale, and search the live web — all from natural language prompts.

Prerequisites

1. Nimble CLI — Required by the nimble-web-tools skill:

npm install -g @nimble-way/nimble-cli

2. Nimble API Key — Get yours from Account Settings and set it as an environment variable:

export NIMBLE_API_KEY="your-api-key"

Or persist it in ~/.claude/settings.json for Claude Code:

{
  "env": {
    "NIMBLE_API_KEY": "your-api-key"
  }
}

Installation

Claude Code
Cursor
Codex
Cortex Code
Vercel CLI

Install from the Nimble Marketplace — this adds both skills and configures the MCP server automatically:

claude plugin marketplace add Nimbleway/agent-skills &&
claude plugin install nimble@nimble-plugin-marketplace

Or load directly from a local clone:

git clone https://github.com/Nimbleway/agent-skills.git &&
claude --plugin-dir /path/to/agent-skills

Step 1: Add the Nimble MCP server to Cursor:

Replace NIMBLE_API_KEY in Cursor Settings > MCP Servers with your actual key.Step 2: Install skills and rules:

npx skills add Nimbleway/agent-skills -a cursor

Step 1: Add the Nimble MCP server to Codex.Codex stores MCP configuration in ~/.codex/config.toml (or a project-scoped .codex/config.toml). Open the file and add:

[mcp_servers.nimble-mcp-server]
url = "https://mcp.nimbleway.com/mcp"
bearer_token_env_var = "NIMBLE_API_KEY"

bearer_token_env_var tells Codex to read your API key from the NIMBLE_API_KEY environment variable at runtime and send it as a bearer token — never hardcode the key in config.toml.

Verify the server is registered from the Codex TUI:

/mcp

Step 2: Install the Nimble skills:

npx skills add Nimbleway/agent-skills -a codex

This installs the nimble-web-tools and nimble-agents skills into Codex’s skill discovery paths (~/.agents/skills/). Codex picks them up automatically; if they don’t appear, restart Codex.Step 3: Verify the skills loaded by running /skills inside Codex, or by typing $ to see the skill mention menu — you should see nimble-web-tools and nimble-agents listed.

Step 1: Add the Nimble MCP server to Cortex Code CLI:

cortex mcp add nimble-mcp-server https://mcp.nimbleway.com/mcp \
  --transport http \
  --header "Authorization: Bearer ${NIMBLE_API_KEY}"

Verify the server is reachable:

cortex mcp start nimble-mcp-server

Cortex Code stores MCP configuration in ~/.snowflake/cortex/mcp.json. You can edit it directly instead of using the CLI — just use ${NIMBLE_API_KEY} syntax for environment variables, never hardcode the key.

Step 2: Install the skills from the Nimble repository:

cortex skill add https://github.com/Nimbleway/agent-skills.git

Or install from inside a Cortex Code CLI session:

/skill add https://github.com/Nimbleway/agent-skills.git

Step 3: Verify the skills loaded:

/skill list

You should see nimble-web-tools and nimble-agents in the listing. Use /skill sync to pull updates from the remote repo later.

npx skills add Nimbleway/agent-skills

The nimble-agents skill requires the MCP server. Connect it manually after installing:

claude mcp add --transport http nimble-mcp-server https://mcp.nimbleway.com/mcp \
  --header "Authorization: Bearer ${NIMBLE_API_KEY}"

What’s Included

Agent Skills are plug-and-play extensions that give AI coding assistants direct access to Nimble’s web data tools. Install once and your AI can search the live web, extract pages, map sites, and run structured data agents — automatically, from natural language.

Skill	Description
nimble-web-tools	Real-time web intelligence — search, extract, map, and crawl via the Nimble CLI
nimble-agents	Find, generate, and run agents to extract structured data from any website
MCP Server	Pre-configured Nimble MCP server connection for agent-to-API access

nimble-web-tools Skill

The nimble-web-tools skill activates automatically when you ask your AI assistant to search the web, extract a page, map a site, or crawl for content.

What it can do

Tool	What it does
Search	Real-time web search with 8 focus modes: general, coding, news, academic, shopping, social, geo, location
Extract	Get clean HTML or markdown from any URL — supports JS rendering and stealth unblocking
Map	Discover all URLs in a domain or sitemap — useful for planning extraction workflows
Crawl	Extract content from an entire website in one request

Example prompts

"Search for recent AI developments"
  → nimble search --query "recent AI developments" --search-depth lite

"Extract the content from this URL"
  → nimble extract --url "https://example.com" --parse --format markdown

"Map all pages on this docs site"
  → nimble map --url "https://docs.example.com" --limit 100

"Crawl the API reference section"
  → nimble crawl run --url "https://docs.example.com/api" --limit 50

Search

# Lite search (default — fast, token-efficient)
nimble search --query "React hooks tutorial" --topic coding --search-depth lite

# Search with AI-generated answer
nimble search --query "what is WebAssembly" --include-answer --search-depth lite

# News with time filter
nimble search --query "AI developments" --topic news --time-range week --search-depth lite

# Filter to specific domains
nimble search --query "auth best practices" \
  --include-domain github.com \
  --include-domain stackoverflow.com \
  --search-depth lite

Use --search-depth lite (default) for fast responses (1–3s). Use --search-depth deep when you need full page content for archiving or full-text analysis.

Extract

# Standard extraction — always pass --parse --format markdown for LLM-readable output
nimble extract --url "https://example.com/article" --parse --format markdown

# With JavaScript rendering (for SPAs and dynamic pages)
nimble extract --url "https://example.com/app" --render --parse --format markdown

# With geo-targeting
nimble extract --url "https://example.com" --country US --city "New York" --parse --format markdown

Map

# Discover all URLs on a site
nimble map --url "https://docs.example.com" --limit 100

# Include subdomains
nimble map --url "https://example.com" --domain-filter subdomains

Crawl

# Crawl a site (always set --limit)
nimble crawl run --url "https://docs.example.com" --limit 50

# Filter to specific paths
nimble crawl run --url "https://example.com" \
  --include-path "/docs" \
  --include-path "/api" \
  --limit 100

# Check crawl status
nimble crawl status --id "crawl-id"

For LLM-friendly output, prefer map + extract --parse --format markdown on individual pages rather than crawl — crawl returns raw HTML which can be very large.

nimble-agents Skill

The nimble-agents skill gives your AI assistant the power to extract structured data from any website. It uses a find-or-generate workflow: search for an existing agent, run it, and get structured data. If no agent exists for a site, it generates a custom one automatically.

How it works

Find an existing agent matching your target website using nimble_agents_list

Inspect

Review the agent’s input/output schema with nimble_agents_get

Run

Execute the agent and get clean, structured results via nimble_agents_run

Generate

If no existing agent fits, create a custom one with nimble_agents_generate

Publish

Save generated agents for future reuse with nimble_agents_publish

Usage

Claude Code
Cursor
Cortex Code
Codex

Use the slash command:

/nimble:nimble-agents extract product details from this Amazon page: https://www.amazon.com/dp/B0DGHRT7PS

Or just describe what you need in plain language — the skill activates automatically when relevant.

Reference the skill in Cursor Agent chat:

/nimble-agents extract product details from this Amazon page: https://www.amazon.com/dp/B0DGHRT7PS

Invoke the skill explicitly from a Cortex Code CLI session:

$nimble-agents extract product details from this Amazon page: https://www.amazon.com/dp/B0DGHRT7PS

Or describe the task in natural language — Cortex Code automatically loads the skill when your prompt matches its trigger keywords (web extraction, scraping, structured data from a URL, etc.).

Invoke the skill explicitly from a Codex CLI, IDE, or app session by mentioning it with $:

$nimble-agents extract product details from this Amazon page: https://www.amazon.com/dp/B0DGHRT7PS

Or describe the task in natural language — Codex will implicitly invoke the skill when your prompt matches its description (web extraction, scraping, structured data from a URL, etc.).

MCP Tools

Tool	Purpose
`nimble_agents_list`	Browse agents by keyword
`nimble_agents_get`	Get agent details and input/output schema
`nimble_agents_generate`	Create a custom agent via natural language
`nimble_agents_run`	Execute an agent and get structured results
`nimble_agents_publish`	Save a generated agent for reuse

Source Code

agent-skills on GitHub

Plugin source, skill definitions, and SKILL.md references

Plugin Installation

Detailed platform-specific installation guide

Web Tools Skill

Full command reference for nimble-web-tools

Agents Skill

Full guide for the nimble-agents skill

​Prerequisites

​Installation

​What’s Included

​nimble-web-tools Skill

​What it can do

​Example prompts

​Search

​Extract

​Map

​Crawl

​nimble-agents Skill

​How it works

​Usage

​MCP Tools

​Source Code

agent-skills on GitHub

Plugin Installation

Web Tools Skill

Agents Skill

Prerequisites

Installation

What’s Included

nimble-web-tools Skill

What it can do

Example prompts

Search

Extract

Map

Crawl

nimble-agents Skill

How it works

Usage

MCP Tools

Source Code