### LLM Context API Endpoint Examples

Source: https://github.com/brave/brave-search-skills/blob/main/skills/llm-context/SKILL.md

Demonstrates how to call the LLM Context API using both GET and POST requests. Ensure you replace ${BRAVE_SEARCH_API_KEY} with your actual API key.

```http
GET  https://api.search.brave.com/res/v1/llm/context
POST https://api.search.brave.com/res/v1/llm/context
```

```bash
curl -s "https://api.search.brave.com/res/v1/llm/context?q=tallest+mountains+in+the+world" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"
```

```bash
curl -s --compressed -X POST "https://api.search.brave.com/res/v1/llm/context" \
  -H "Accept: application/json" \
  -H "Accept-Encoding: gzip" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{"q": "tallest mountains in the world"}'
```

--------------------------------

### Install bx CLI (macOS/Linux)

Source: https://github.com/brave/brave-search-skills/blob/main/skills/bx/SKILL.md

Installs the Brave Search CLI using a script. Ensure you have curl installed.

```bash
# Install (macOS/Linux)
curl -fsSL https://raw.githubusercontent.com/brave/brave-search-cli/main/scripts/install.sh | sh
```

--------------------------------

### Install Brave Search Skills for Windsurf (Project-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills to the project's .windsurf/skills directory using curl.

```bash
# Project-level
mkdir -p .windsurf/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C .windsurf/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Install Brave Search Skills for Windsurf (User-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills to the user's ~/.codeium/windsurf/skills directory using curl.

```bash
# User-level
mkdir -p ~/.codeium/windsurf/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C ~/.codeium/windsurf/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Install Brave Search Skills for OpenClaw (curl)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills to the user's ~/.openclaw/skills directory using curl.

```bash
mkdir -p ~/.openclaw/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C ~/.openclaw/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Install Brave Search Skills for Claude Code Manually

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Install Brave Search skills for Claude Code by cloning the repository and copying the skills directory. This method provides manual control over the installation.

```bash
git clone https://github.com/brave/brave-search-skills
cp -r brave-search-skills/skills/* ~/.claude/skills/   # user-level
cp -r brave-search-skills/skills/* .claude/skills/      # project-level
```

--------------------------------

### Install Brave Search Skills for GitHub Copilot (curl)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills to the .github/skills directory using curl. This is for GitHub Copilot integration.

```bash
mkdir -p .github/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C .github/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Install Brave Search Skills using OpenSkills

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills using the third-party universal installer 'openskills'. This command overwrites existing skills.

```bash
npx openskills install brave/brave-search-skills
```

--------------------------------

### Install Brave Search Skills for User-Level (Cursor)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills to the user's Cursor skills directory using curl. Ensure the directory exists before running.

```bash
mkdir -p ~/.cursor/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C ~/.cursor/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Install Brave Search CLI on macOS/Linux

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Installs the Brave Search CLI using a curl script. Ensure you have curl installed.

```bash
curl -fsSL https://raw.githubusercontent.com/brave/brave-search-cli/main/scripts/install.sh | sh
```

--------------------------------

### Install Brave Search Skills for Codex (Project-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills to the project's .agents/skills directory using curl. This is for Codex integration.

```bash
# Project-level
mkdir -p .agents/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C .agents/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Install Brave Search Skills for Codex (User-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills to the user's ~/.agents/skills directory using curl. This is for Codex integration.

```bash
# User-level
mkdir -p ~/.agents/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C ~/.agents/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Install Brave Search CLI on Windows (PowerShell)

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Installs the Brave Search CLI using a PowerShell command. Requires PowerShell execution policy to be set.

```powershell
powershell -ExecutionPolicy Bypass -c "irm https://raw.githubusercontent.com/brave/brave-search-cli/main/scripts/install.ps1 | iex"
```

--------------------------------

### Install Brave Search Skills for Cursor via curl (Project-level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Install Brave Search skills for Cursor at the project level using curl to download and extract the archive. This method is suitable for project-specific configurations.

```bash
# Project-level
mkdir -p .cursor/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C .cursor/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Checking for breaking changes before upgrading

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Use `context` to find migration guides and `news` to check for release information. Specify the library/framework and version.

```bash
bx context "Next.js 15 breaking changes migration guide" --max-tokens 8192
bx news "Next.js 15 release" --freshness pm
```

--------------------------------

### Install Brave Search Skills for Claude Code via Plugin Marketplace

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Install the Brave Search skills for Claude Code using the plugin marketplace. This method ensures automatic updates and does not require a git clone.

```bash
/plugin marketplace add brave/brave-search-skills
/plugin install brave-search-skills@brave-search
```

--------------------------------

### Install Brave Search Skills for Other Agents (curl)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Installs Brave Search Skills to a specified skills directory using curl. Adjust '<skills-dir>' to your agent's directory.

```bash
mkdir -p <skills-dir> && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C <skills-dir> --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Get search suggestions

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Use the `suggest` command to get search query suggestions. Pipe the output to `jq` to extract the suggested queries.

```bash
bx suggest "how to implement" --count 10 | jq '.results[].query'
```

--------------------------------

### Install Brave Search Skills for Claude Code via curl (Project-level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Install Brave Search skills for Claude Code at the project level using curl to download and extract the archive. This method avoids git and the marketplace.

```bash
# Project-level
mkdir -p .claude/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C .claude/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Goggles: Hosted Example

Source: https://github.com/brave/brave-search-skills/blob/main/skills/llm-context/SKILL.md

Use hosted Goggles by providing a URL to a Goggle rules file on GitHub or GitLab. The file must be registered.

```bash
--data-urlencode "goggles=https://raw.githubusercontent.com/brave/goggles-quickstart/main/goggles/1k_short.goggle"
```

--------------------------------

### Install Brave Search Skills for Claude Code via curl (User-level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Install Brave Search skills for Claude Code at the user level using curl to download and extract the archive. This method avoids git and the marketplace.

```bash
# User-level (available in all projects)
mkdir -p ~/.claude/skills && curl -sL https://github.com/brave/brave-search-skills/archive/main.tar.gz | tar xz -C ~/.claude/skills --strip-components=2 brave-search-skills-main/skills
```

--------------------------------

### Rich Response Example

Source: https://github.com/brave/brave-search-skills/blob/main/skills/suggest/SKILL.md

This JSON example shows the structure of a rich suggestion response when 'rich=true' is specified. It includes entity information like title, description, and image.

```json
{
  "type": "suggest",
  "query": { "original": "albert" },
  "results": [
    {
      "query": "albert einstein",
      "is_entity": true,
      "title": "Albert Einstein",
      "description": "German-born theoretical physicist",
      "img": "https://imgs.search.brave.com/..."
    },
    { "query": "albert einstein quotes", "is_entity": false }
  ]
}
```

--------------------------------

### Rich Data Enrichment Workflow

Source: https://github.com/brave/brave-search-skills/blob/main/skills/web-search/SKILL.md

This example demonstrates how to enable rich callback for search queries and then retrieve rich data using the provided callback key.

```bash
# 1. Search with rich callback enabled
curl -s "https://api.search.brave.com/res/v1/web/search?q=weather+san+francisco&enable_rich_callback=true" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"

# Response includes: "rich": {"hint": {"callback_key": "abc123...", "vertical": "weather"}}

# 2. Get rich data with the callback key
curl -s "https://api.search.brave.com/res/v1/web/rich?callback_key=abc123..." \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"
```

--------------------------------

### Brave Search CLI Error Output Example

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Example of the JSON-formatted error output from the Brave Search CLI, indicating a rate limiting issue.

```json
error: rate limited (429) — Request rate limit exceeded for plan.
hint: retry after a short delay, or upgrade plan for higher rate limits
{"type":"ErrorResponse","error":{"code":"RATE_LIMITED","status":429,...}}
```

--------------------------------

### Direct AI answer (streaming)

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Use the `answers` command to get a direct AI-generated explanation. It streams responses by default.

```bash
bx answers "explain Rust lifetimes with examples"
```

--------------------------------

### Evaluating a dependency

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Use `context` to check for security issues or maintenance status of a crate, and `news` to get fresh information. Specify the crate name and relevant keywords.

```bash
bx context "reqwest crate security issues maintained 2026" --threshold strict
bx news "reqwest Rust crate" --freshness pm
```

--------------------------------

### Goggles: Inline Example

Source: https://github.com/brave/brave-search-skills/blob/main/skills/llm-context/SKILL.md

Use inline Goggles rules for custom ranking without registration. Rules are combined with commas and separated by \n.

```bash
--data-urlencode 'goggles=$discard\n$site=example.com'
```

--------------------------------

### Basic Suggestions (cURL)

Source: https://github.com/brave/brave-search-skills/blob/main/skills/suggest/SKILL.md

Use this snippet to get basic query suggestions by providing a query parameter 'q'. Ensure your API key is set in the BRAVE_SEARCH_API_KEY environment variable.

```bash
curl -s "https://api.search.brave.com/res/v1/suggest/search?q=how+to+" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"
```

--------------------------------

### Direct Comparison Results (Example)

Source: https://github.com/brave/brave-search-skills/wiki/Evaluation-of-major-AI-powered-answer-engines

Displays the percentage breakdown of direct comparison results between AI answer engines. This format is used to show relative performance in head-to-head evaluations.

```text
BOTH EQUAL      │████████████████████████████████████████│ 38.73%
ChatGPT         │███████████████████████████│ 26.87%
AI Mode         │███████████████████████████████████│ 34.40%
```

```text
BOTH EQUAL      │█████████████████│ 27.73%
ChatGPT         │█████│ 9.07%
Grok            │████████████████████████████████████████│ 63.20%
```

```text
BOTH EQUAL      │████████████████████████████│ 34.07%
ChatGPT         │████████████████████████████████████████│ 48.00%
Perplexity      │██████████████│ 17.93%
```

```text
BOTH EQUAL      │███████████████████│ 30.07%
Grok            │████████████████████████████████████████│ 60.53%
AI Mode         │██████│ 9.40%
```

```text
BOTH EQUAL      │█████│ 15.47%
Grok            │████████████████████████████████████████│ 79.33%
Perplexity      │██│ 5.20%
```

--------------------------------

### Configure direnv for API Key

Source: https://github.com/brave/brave-search-skills/blob/main/docs/api-key-setup.md

Use direnv to automatically load your Brave Search API key for directory-scoped configurations. Ensure direnv is installed and the .envrc file is allowed.

```bash
# Install direnv (https://direnv.net), then in your project directory:
echo 'export BRAVE_SEARCH_API_KEY="your-api-key-here"' >> .envrc
direnv allow
```

--------------------------------

### Spellcheck API Quick Start (cURL)

Source: https://github.com/brave/brave-search-skills/blob/main/skills/spellcheck/SKILL.md

This cURL command demonstrates how to call the spellcheck endpoint with a misspelled query and language/country parameters. Ensure you have set the BRAVE_SEARCH_API_KEY environment variable.

```bash
curl -s "https://api.search.brave.com/res/v1/spellcheck/search" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}" \
  -G \
  --data-urlencode "q=artifical inteligence" \
  --data-urlencode "lang=en" \
  --data-urlencode "country=US"
```

--------------------------------

### LLM Context API with Goggles

Source: https://github.com/brave/brave-search-skills/blob/main/skills/llm-context/SKILL.md

Shows how to use the LLM Context API with Goggles for custom re-ranking of search results. This example filters results to specific sites.

```bash
curl -s "https://api.search.brave.com/res/v1/llm/context" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}" \
  -G \
  --data-urlencode "q=rust programming" \
  --data-urlencode 'goggles=$discard
$site=docs.rs
$site=rust-lang.org'
```

--------------------------------

### Python: Streaming with Citations (Single-Search) OpenAI SDK Usage

Source: https://github.com/brave/brave-search-skills/blob/main/skills/answers/SKILL.md

This example demonstrates how to receive search results as a stream, with inline citation tags. Set `stream=True` and `enable_citations=True` in `extra_body`.

```python
from openai import OpenAI

client = OpenAI(
    base_url="https://api.search.brave.com/res/v1",
    api_key="your-brave-api-key",
)

stream = client.chat.completions.create(
    model="brave",
    messages=[{"role": "user", "content": "What are the current trends in renewable energy?"}],
    stream=True,
    extra_body={"enable_citations": True}
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")
```

--------------------------------

### Answers (AI-Grounded) - Fast Single Search

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Get AI-grounded answers using the OpenAI SDK-compatible endpoint. This example performs a fast, blocking single search. Requires an API key and JSON payload.

```bash
curl -X POST "https://api.search.brave.com/res/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}" \
  -d '{
    "messages": [{"role": "user", "content": "What programming languages are trending in 2026?"}],
    "model": "brave",
    "stream": false
  }'
```

--------------------------------

### Example Response Format

Source: https://github.com/brave/brave-search-skills/blob/main/skills/local-descriptions/SKILL.md

This JSON structure shows how descriptions are returned, including the POI ID and the AI-generated markdown description. Note that the 'description' field can be null if no description is available for a POI.

```json
{
  "type": "local_descriptions",
  "results": [
    {
      "type": "local_description",
      "id": "loc4CQWMJWLD4VBEBZ62XQLJTGK6YCJEEJDNAAAAAAA=",
      "description": "### Overview\nA cozy neighborhood cafe known for its **artisanal coffee**..."
    }
  ]
}
```

--------------------------------

### Getting POI IDs using Web Search API (cURL)

Source: https://github.com/brave/brave-search-skills/blob/main/skills/local-descriptions/SKILL.md

This example demonstrates how to first obtain POI IDs using the Web Search API with the `result_filter=locations` parameter. These IDs are then used to fetch descriptions from the Local Descriptions API.

```bash
# 1. Search for local businesses
curl -s "https://api.search.brave.com/res/v1/web/search?q=restaurants+san+francisco&result_filter=locations" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"

# 2. Extract POI IDs from locations.results[].id
# 3. Use those IDs with local/pois and local/descriptions
```

--------------------------------

### Control Token Budget for Context Search

Source: https://github.com/brave/brave-search-skills/blob/main/skills/bx/SKILL.md

Example of using flags to control the token budget for the default context search command, adjusting maximum tokens, tokens per URL, max URLs, and relevance threshold.

```bash
bx "topic" --max-tokens 4096 --max-tokens-per-url 1024 --max-urls 5 --threshold strict
```

--------------------------------

### Display bx CLI help information

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Shows all available commands and flags for the Brave Search CLI. Use `bx <command> --help` for specific command details.

```bash
bx --help
```

--------------------------------

### Example Local POI Response

Source: https://github.com/brave/brave-search-skills/blob/main/skills/local-pois/SKILL.md

This is an example of a JSON response when querying for local points of interest. It includes detailed information about a business, such as its address, contact information, rating, and opening hours.

```json
{
  "type": "local_pois",
  "results": [
    {
      "type": "location_result",
      "title": "Park Mediterranean Grill",
      "url": "https://yelp.com/biz/park-mediterranean-grill-sf",
      "provider_url": "https://yelp.com/biz/park-mediterranean-grill-sf",
      "id": "loc4CQWMJWLD4VBEBZ62XQLJTGK6YCJEEJDNAAAAAAA=",
      "postal_address": {
        "type": "PostalAddress",
        "displayAddress": "123 Main St, San Francisco, CA 94102",
        "streetAddress": "123 Main St",
        "addressLocality": "San Francisco",
        "addressRegion": "CA",
        "postalCode": "94102",
        "country": "US"
      },
      "contact": { "telephone": "+1 415-555-0123" },
      "thumbnail": {
        "src": "https://example.com/thumb.jpg",
        "original": "https://example.com/original.jpg"
      },
      "rating": {
        "ratingValue": 4.5,
        "bestRating": 5.0,
        "reviewCount": 234
      },
      "opening_hours": {
        "current_day": [
          { "abbr_name": "Mon", "full_name": "Monday", "opens": "07:00", "closes": "21:00" }
        ]
      },
      "coordinates": [37.7749, -122.4194],
      "distance": { "value": 0.3, "units": "miles" },
      "categories": ["Mediterranean", "Greek"],
      "price_range": "$$",
      "serves_cuisine": ["Mediterranean", "Greek"],
      "timezone": "America/Los_Angeles"
    }
  ]
}
```

--------------------------------

### Use bx with API Key Flag

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Demonstrates using the `--api-key` flag to specify the API key for a web search. This method is less secure as the key may be visible in process listings.

```bash
bx --api-key KEY web "test"
```

--------------------------------

### Suggestions with All Parameters (cURL)

Source: https://github.com/brave/brave-search-skills/blob/main/skills/suggest/SKILL.md

This snippet demonstrates how to use all available parameters for the Suggest API, including query, language, country, count, and rich suggestions. It's useful for fine-tuning suggestion results.

```bash
curl -s "https://api.search.brave.com/res/v1/suggest/search" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}" \
  -G \
  --data-urlencode "q=albert" \
  --data-urlencode "country=US" \
  --data-urlencode "lang=en" \
  --data-urlencode "count=10" \
  --data-urlencode "rich=true"
```

--------------------------------

### Using Hosted Goggles with Brave Search CLI

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Reference a hosted `.goggle` file from GitHub or GitLab directly in the CLI to filter search results.

```bash
bx web "distributed systems" \
  --goggles 'https://raw.githubusercontent.com/brave/goggles-quickstart/main/goggles/tech_blogs.goggle'
```

--------------------------------

### Spellcheck Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Performs spell correction on search queries for cleanup. Supports GET requests.

```APIDOC
## GET /res/v1/spellcheck/search

### Description
Corrects spelling errors in search queries to improve relevance.

### Method
GET

### Endpoint
/res/v1/spellcheck/search

### Parameters
#### Query Parameters
- **q** (string) - Required - The search query to spellcheck.

### Response
#### Success Response (200)
- **spellcheck** (object) - Contains spell correction suggestions.
```

--------------------------------

### Suggest Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Provides query autocomplete suggestions with a response time under 100ms. Supports GET requests.

```APIDOC
## GET /res/v1/suggest/search

### Description
Offers query autocomplete suggestions, designed for fast responses.

### Method
GET

### Endpoint
/res/v1/suggest/search

### Parameters
#### Query Parameters
- **q** (string) - Required - The partial search query for autocomplete.

### Response
#### Success Response (200)
- **suggestions** (array) - An array of suggested search queries.
```

--------------------------------

### News Search Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

News article search with freshness filtering capabilities. Supports GET requests.

```APIDOC
## GET /res/v1/news/search

### Description
Retrieves news articles, allowing filtering by freshness.

### Method
GET

### Endpoint
/res/v1/news/search

### Parameters
#### Query Parameters
- **q** (string) - Required - The search query for news.

### Response
#### Success Response (200)
- **news** (array) - An array of news article objects.
```

--------------------------------

### Copy Brave Search Skills Manually for Windsurf (User-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Copies Brave Search Skills from a git clone to the user's ~/.codeium/windsurf/skills directory. Requires a prior git clone.

```bash
cp -r brave-search-skills/skills/* ~/.codeium/windsurf/skills/
```

--------------------------------

### LLM Context Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Pre-extracted web content for LLM grounding. Supports GET and POST requests.

```APIDOC
## GET /res/v1/llm/context

### Description
Returns search results with pre-extracted web content, optimized for LLM grounding.

### Method
GET

### Endpoint
/res/v1/llm/context

### Parameters
#### Query Parameters
- **q** (string) - Required - The search query.
- **goggles** (string) - Optional - Custom ranking instructions.

### Request Example
```bash
curl -X GET "https://api.search.brave.com/res/v1/llm/context?q=search+API+for+grounding+LLMs" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"
```

### Response
#### Success Response (200)
- **results** (array) - Search results with pre-extracted web content.
```

--------------------------------

### Copy Brave Search Skills Manually for Windsurf (Project-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Copies Brave Search Skills from a git clone to the project's .windsurf/skills directory. Requires a prior git clone.

```bash
cp -r brave-search-skills/skills/* .windsurf/skills/
```

--------------------------------

### Images Search Endpoint

Source: https://github.com/brave/brave-search-skills/blob/main/skills/images-search/SKILL.md

The GET endpoint for performing image searches via the Brave Search API.

```http
GET https://api.search.brave.com/res/v1/images/search
```

--------------------------------

### Perform a basic web search with bx

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Executes a web search query using the Brave Search CLI. This is the default behavior for bare queries.

```bash
bx "your search query"
```

--------------------------------

### Get AI-Generated Answers

Source: https://github.com/brave/brave-search-skills/blob/main/skills/bx/SKILL.md

Retrieves synthesized explanations for a query. This command streams results by default and cites sources.

```bash
bx answers "query"
```

--------------------------------

### Videos Search Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Video search functionality that includes filtering by duration, views, and creator. Supports GET requests.

```APIDOC
## GET /res/v1/videos/search

### Description
Searches for video content, with options to filter by duration, views, and creator.

### Method
GET

### Endpoint
/res/v1/videos/search

### Parameters
#### Query Parameters
- **q** (string) - Required - The search query for videos.

### Response
#### Success Response (200)
- **videos** (array) - An array of video objects.
```

--------------------------------

### Basic Video Search with cURL

Source: https://github.com/brave/brave-search-skills/blob/main/skills/videos-search/SKILL.md

Performs a basic video search for 'python tutorial'. Ensure your API key is set in the BRAVE_SEARCH_API_KEY environment variable.

```bash
curl -s "https://api.search.brave.com/res/v1/videos/search?q=python+tutorial" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"
```

--------------------------------

### Local Descriptions Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Generates AI-powered text descriptions for Points of Interest (POIs) based on their IDs. Supports GET requests.

```APIDOC
## GET /res/v1/local/descriptions

### Description
Provides AI-generated text descriptions for POIs using their IDs.

### Method
GET

### Endpoint
/res/v1/local/descriptions

### Parameters
#### Query Parameters
- **id** (string) - Required - The ID of the POI for which to generate a description.

### Response
#### Success Response (200)
- **local_descriptions** (array) - An array of generated POI description objects.
```

--------------------------------

### Images Search Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Image search functionality returning image thumbnails, supporting up to 200 results. Supports GET requests.

```APIDOC
## GET /res/v1/images/search

### Description
Performs image search and returns thumbnails for up to 200 results.

### Method
GET

### Endpoint
/res/v1/images/search

### Parameters
#### Query Parameters
- **q** (string) - Required - The search query for images.

### Response
#### Success Response (200)
- **images** (array) - An array of image objects, including thumbnails.
```

--------------------------------

### Goggles DSL Quick Reference

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

A reference table for common Goggles DSL rules including boosting, downranking, discarding domains, and path-based targeting with wildcards.

```markdown
| Rule | Effect | Example |
|---|---|---|
| `$boost=N,site=DOMAIN` | Promote domain (N=1-10) | `$boost=3,site=docs.rs` |
| `$downrank=N,site=DOMAIN` | Demote domain (N=1-10) | `$downrank=5,site=medium.com` |
| `$discard,site=DOMAIN` | Remove domain entirely | `$discard,site=w3schools.com` |
| `/path/$boost=N` | Boost matching URL paths | `/docs/$boost=5` |
| `*pattern*$boost=N` | Wildcard URL matching | `*api*$boost=3` |
| Generic `$discard` | Allowlist mode — discard all unmatched | `$discard` (as first rule) |
```

--------------------------------

### Web Search Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Standard web search providing ranked results with snippets and rich data. Supports GET requests.

```APIDOC
## GET /res/v1/web/search

### Description
Provides ranked web results with snippets and metadata.

### Method
GET

### Endpoint
/res/v1/web/search

### Parameters
#### Query Parameters
- **q** (string) - Required - The search query.

### Request Example
```bash
curl -s "https://api.search.brave.com/res/v1/web/search?q=retrieval+augmented+generation+explained" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"
```

### Response
#### Success Response (200)
- **webPages** (object) - Contains search results, snippets, and metadata.
```

--------------------------------

### `bx context` response shape

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Example JSON response shape for the `bx context` command, showing RAG grounding results.

```json
{
  "grounding": {
    "generic": [
      { "url": "...", "title": "...", "snippets": ["extracted content...", "..."] }
    ]
  }
}
```

--------------------------------

### LLM Context Endpoint

Source: https://github.com/brave/brave-search-skills/blob/main/skills/llm-context/SKILL.md

This endpoint retrieves pre-extracted web content optimized for LLM grounding. It can be accessed via GET or POST requests.

```APIDOC
## GET /res/v1/llm/context

### Description
Retrieves pre-extracted web content optimized for LLM grounding. Supports query parameters for search and context control.

### Method
GET

### Endpoint
https://api.search.brave.com/res/v1/llm/context

### Parameters
#### Query Parameters
- **q** (string) - Required - Search query (1-400 chars, max 50 words)
- **country** (string) - Optional - Search country (2-letter country code or `ALL`)
- **search_lang** (string) - Optional - Language preference (2+ char language code)
- **count** (int) - Optional - Max search results to consider (1-50)

#### Context Size Parameters
- **maximum_number_of_urls** (int) - Optional - Max URLs in response (1-50)
- **maximum_number_of_tokens** (int) - Optional - Approximate max tokens in context (1024-32768)
- **maximum_number_of_snippets** (int) - Optional - Max snippets across all URLs (1-100)
- **maximum_number_of_tokens_per_url** (int) - Optional - Max tokens per individual URL (512-8192)
- **maximum_number_of_snippets_per_url** (int) - Optional - Max snippets per individual URL (1-100)

#### Filtering & Local Parameters
- **context_threshold_mode** (string) - Optional - Relevance threshold for including content (`strict`/`balanced`/`lenient`)
- **enable_local** (bool) - Optional - Local recall control (`true`/`false`/`null`)
- **goggles** (string/list) - Optional - Goggle URL or inline definition for custom re-ranking

### Request Example
```bash
curl -s "https://api.search.brave.com/res/v1/llm/context?q=tallest+mountains+in+the+world" \
  -H "Accept: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}"
```

## POST /res/v1/llm/context

### Description
Retrieves pre-extracted web content optimized for LLM grounding using a JSON request body.

### Method
POST

### Endpoint
https://api.search.brave.com/res/v1/llm/context

### Parameters
#### Request Body
- **q** (string) - Required - Search query (1-400 chars, max 50 words)
- **country** (string) - Optional - Search country (2-letter country code or `ALL`)
- **search_lang** (string) - Optional - Language preference (2+ char language code)
- **count** (int) - Optional - Max search results to consider (1-50)
- **maximum_number_of_urls** (int) - Optional - Max URLs in response (1-50)
- **maximum_number_of_tokens** (int) - Optional - Approximate max tokens in context (1024-32768)
- **maximum_number_of_snippets** (int) - Optional - Max snippets across all URLs (1-100)
- **maximum_number_of_tokens_per_url** (int) - Optional - Max tokens per individual URL (512-8192)
- **maximum_number_of_snippets_per_url** (int) - Optional - Max snippets per individual URL (1-100)
- **context_threshold_mode** (string) - Optional - Relevance threshold for including content (`strict`/`balanced`/`lenient`)
- **enable_local** (bool) - Optional - Local recall control (`true`/`false`/`null`)
- **goggles** (string/list) - Optional - Goggle URL or inline definition for custom re-ranking

### Request Example
```bash
curl -s --compressed -X POST "https://api.search.brave.com/res/v1/llm/context" \
  -H "Accept: application/json" \
  -H "Accept-Encoding: gzip" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{"q": "tallest mountains in the world"}'
```

### Response
#### Success Response (200)
- **url** (string) - The URL of the web page.
- **title** (string) - The title of the web page.
- **content** (string) - The extracted text content from the page.
- **snippets** (array) - A list of relevant snippets from the page.
  - **snippet** (string) - A text snippet.
  - **url** (string) - The URL the snippet is from.
- **tables** (array) - A list of extracted tables from the page.
  - **caption** (string) - The caption of the table.
  - **data** (array) - The table data.
    - (array of strings) - A row in the table.
- **code_blocks** (array) - A list of extracted code blocks from the page.
  - **code** (string) - The code block content.
  - **language** (string) - The programming language of the code block.

#### Response Example
```json
{
  "url": "https://example.com/page",
  "title": "Example Page Title",
  "content": "This is the extracted text content.",
  "snippets": [
    {
      "snippet": "This is a relevant snippet.",
      "url": "https://example.com/page"
    }
  ],
  "tables": [
    {
      "caption": "Example Table",
      "data": [
        ["Header 1", "Header 2"],
        ["Row 1 Col 1", "Row 1 Col 2"]
      ]
    }
  ],
  "code_blocks": [
    {
      "code": "print('Hello, world!')",
      "language": "python"
    }
  ]
}
```
```

--------------------------------

### Copy Brave Search Skills Manually (Project-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Copies Brave Search Skills from a git clone to the current directory's skills folder. Requires a prior git clone.

```bash
cp -r brave-search-skills/skills/* .cursor/skills/
```

--------------------------------

### Videos Search Endpoint

Source: https://github.com/brave/brave-search-skills/blob/main/skills/videos-search/SKILL.md

This endpoint allows you to search for videos. You can use either the GET or POST method. POST is recommended for longer queries.

```APIDOC
## Videos Search

### Description
Searches for videos based on a query and returns a list of video results, including title, URL, thumbnail, duration, view count, and creator. Supports various filtering and pagination options.

### Method
GET, POST

### Endpoint
https://api.search.brave.com/res/v1/videos/search

### Parameters
#### Query Parameters
- **q** (string) - Required - Search query (1-400 chars, max 50 words)
- **country** (string) - Optional - Search country (2-letter country code or `ALL`). Default: `US`
- **search_lang** (string) - Optional - Language preference (2+ char language code). Default: `en`
- **ui_lang** (string) - Optional - UI language (e.g., "en-US"). Default: `en-US`
- **count** (int) - Optional - Number of results (1-50). Default: `20`
- **offset** (int) - Optional - Page offset (0-9). Default: `0`
- **safesearch** (string) - Optional - Adult content filter (`off`/`moderate`/`strict`). Default: `moderate`
- **freshness** (string) - Optional - Time filter (`pd`/`pw`/`pm`/`py` or date range `YYYY-MM-DDtoYYYY-MM-DD`)
- **spellcheck** (bool) - Optional - Auto-correct query. Default: `true`
- **operators** (bool) - Optional - Apply search operators. Default: `true`
- **include_fetch_metadata** (bool) - Optional - Include `fetched_content_timestamp` in results. Default: `false`

### Request Example
```json
{
  "q": "python tutorial",
  "country": "US",
  "count": 10,
  "freshness": "pw"
}
```

### Response
#### Success Response (200)
- **type** (string) - Type of the response, expected to be "videos".
- **query** (object) - Information about the search query.
  - **original** (string) - The original search query.
  - **spellcheck_off** (bool) - Indicates if spellcheck was turned off.
- **extra** (object) - Additional information about the search.
  - **might_be_offensive** (bool) - Indicates if the results might contain offensive content.
- **results** (array) - An array of video result objects.
  - **type** (string) - Type of the result, expected to be "video_result".
  - **title** (string) - The title of the video.
  - **url** (string) - The URL of the video.
  - **description** (string) - A brief description of the video.
  - **age** (string) - The age of the video content (e.g., "February 12, 2025").
  - **page_age** (string) - The date the page was created.
  - **page_fetched** (string) - The timestamp when the page was fetched.
  - **thumbnail** (object) - Thumbnail image details.
    - **src** (string) - URL for the thumbnail image.
    - **original** (string) - Original URL for the thumbnail.
  - **video** (object) - Video specific details.
    - **duration** (string) - Duration of the video (e.g., "03:45:00").
    - **views** (int) - Number of views.
    - **creator** (string) - The creator of the video.
    - **publisher** (string) - The publisher of the video.
    - **requires_subscription** (bool) - Whether the video requires a subscription.
    - **tags** (array) - Array of tags associated with the video.
    - **author** (object) - Author details.
      - **name** (string) - Author's name.
      - **url** (string) - Author's URL.
  - **meta_url** (object) - Metadata about the URL.
    - **scheme** (string) - URL scheme (e.g., "https").
    - **netloc** (string) - Network location (e.g., "youtube.com").
    - **hostname** (string) - Hostname of the URL.
    - **favicon** (string) - URL for the favicon.
    - **path** (string) - Path component of the URL.

#### Response Example
```json
{
  "type": "videos",
  "query": {
    "original": "python tutorial",
    "spellcheck_off": false
  },
  "extra": {
    "might_be_offensive": false
  },
  "results": [
    {
      "type": "video_result",
      "title": "Python Tutorial for Beginners",
      "url": "https://www.youtube.com/watch?v=rfscVS0vtbw",
      "description": "Learn Python programming from scratch...",
      "age": "February 12, 2025",
      "page_age": "2025-02-12T00:00:00",
      "page_fetched": "2025-02-12T15:00:00Z",
      "thumbnail": {
        "src": "https://imgs.search.brave.com/...",
        "original": "https://i.ytimg.com/vi/rfscVS0vtbw/hqdefault.jpg"
      },
      "video": {
        "duration": "03:45:00",
        "views": 1523000,
        "creator": "freeCodeCamp",
        "publisher": "YouTube",
        "requires_subscription": false,
        "tags": ["python", "programming"],
        "author": {
          "name": "freeCodeCamp.org",
          "url": "https://www.youtube.com/@freecodecamp"
        }
      },
      "meta_url": {
        "scheme": "https",
        "netloc": "youtube.com",
        "hostname": "www.youtube.com",
        "favicon": "https://imgs.search.brave.com/...",
        "path": "\u203a watch"
      }
    }
  ]
}
```
```

--------------------------------

### Inline Goggles Rules for Boosting and Demoting

Source: https://github.com/brave/brave-search-skills/blob/main/skills/bx/SKILL.md

Apply inline Goggles rules to boost or demote specific sites and paths, or to enable allowlist mode.

```bash
# Boost official docs, demote blog posts
bx "axum middleware tower" \
  --goggles '$boost=5,site=docs.rs
$boost=3,site=github.com
/docs/$boost=5
/blog/$downrank=3' --max-tokens 4096

# Allowlist mode — only include matched sites
bx "Python asyncio patterns" \
  --goggles '$discard
$boost,site=docs.python.org
$boost,site=peps.python.org'
```

--------------------------------

### Copy Brave Search Skills Manually (User-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Copies Brave Search Skills from a git clone to the user's skills directory. Requires a prior git clone.

```bash
cp -r brave-search-skills/skills/* ~/.cursor/skills/
```

--------------------------------

### Blocking Single-Search cURL Example

Source: https://github.com/brave/brave-search-skills/blob/main/skills/answers/SKILL.md

Use this snippet for a fast, AI-grounded answer from a single search without streaming. Ensure your BRAVE_SEARCH_API_KEY is set.

```bash
curl -X POST "https://api.search.brave.com/res/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "X-Subscription-Token: ${BRAVE_SEARCH_API_KEY}" \
  -d '{
    "messages": [{"role": "user", "content": "How does the James Webb Space Telescope work?"}],
    "model": "brave",
    "stream": false
  }'
```

--------------------------------

### `bx answers` streaming response shape

Source: https://github.com/brave/brave-search-skills/blob/main/clawhub/bx-search/SKILL.md

Example JSON response shape for the `bx answers` command when streaming, showing incremental content chunks.

```json
{"choices": [{"delta": {"content": "R"}}]}
{"choices": [{"delta": {"content": "u"}}]}
{"choices": [{"delta": {"content": "s"}}]}
{"choices": [{"delta": {"content": "t"}}]}
{"choices": [{"delta": {"content": " "}}]}
```

--------------------------------

### Corrective RAG Loop with Brave Search

Source: https://github.com/brave/brave-search-skills/blob/main/skills/bx/SKILL.md

Demonstrates a multi-step search process for complex queries, starting broad and narrowing down, then requesting a synthesized answer.

```bash
# 1. Broad search
bx "axum middleware authentication" --max-tokens 4096
# 2. Too general? Narrow with strict threshold
bx "axum middleware tower layer authentication example" --threshold strict --max-tokens 4096
# 3. Still need synthesis? Ask for an answer
bx answers "how to implement JWT auth middleware in axum" --enable-research
```

--------------------------------

### Local POIs Skill

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Retrieves details for local businesses and Points of Interest (POIs), including ratings, hours, and contact information. Supports GET requests.

```APIDOC
## GET /res/v1/local/pois

### Description
Fetches detailed information for local businesses and POIs, such as ratings, hours, and contact details.

### Method
GET

### Endpoint
/res/v1/local/pois

### Parameters
#### Query Parameters
- **id** (string) - Required - The ID of the local business or POI.

### Response
#### Success Response (200)
- **local_pois** (array) - An array of local POI objects with details.
```

--------------------------------

### Copy Brave Search Skills Manually for GitHub Copilot

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Copies Brave Search Skills from a git clone to the .github/skills directory. Requires a prior git clone.

```bash
cp -r brave-search-skills/skills/* .github/skills/
```

--------------------------------

### Copy Brave Search Skills Manually for Codex (User-Level)

Source: https://github.com/brave/brave-search-skills/blob/main/README.md

Copies Brave Search Skills from a git clone to the user's ~/.agents/skills directory. Requires a prior git clone.

```bash
cp -r brave-search-skills/skills/* ~/.agents/skills/
```