AgentPatch provides context-optimized APIs for AI agents, with microtransaction billing.
The simplest integration is `pip install agentpatch` — a zero-dependency CLI and Python SDK.

AI agents can discover and invoke APIs without requiring additional accounts to be configured
by humans. Every response includes structured schemas and examples optimized for LLM consumption.

Supported tools include email sending/receiving, web search/scraping, and more.
New APIs are added regularly.

## Available Tools

### Amazon Search
- Description: Search Amazon products. Returns product listings with prices, ratings, and ASINs.
- Path: /api/tools/amazon-search
- Price: 400 credits ($0.04)
- Input: {"query": "string (required) — Product search query", "country": "string — Amazon domain (e.g., 'us', 'uk', 'de', 'jp')", "page": "integer — Page number"}
- Output: {"products": "array", "query": "string", "country": "string"}

### Bing Search
- Description: Search Bing and get structured results. Returns search results with titles, links, and snippets.
- Path: /api/tools/bing-search
- Price: 50 credits ($0.01)
- Input: {"query": "string (required) — Search query", "country": "string — Country code (e.g., 'us', 'uk')", "count": "integer — Number of results (default: 10)", "page": "integer — Page number (0-indexed)", "safe_search": "Off|Moderate|Strict — Safe search level (default: Moderate)"}
- Output: {"results": "array", "query": "string"}

### Census Data
- Description: Get demographic and economic data from the US Census Bureau American Community Survey. Supports population, income, housing, education, and more at national, state, county, ZIP, and place levels.
- Path: /api/tools/census-data
- Price: 75 credits ($0.01)
- Input: {"variables": "array (required) — Human-readable variable names: total_population, median_age, median_household_income, per_capita_income, poverty_rate, unemployment_rate, bachelors_degree_or_higher, total_housing_units, median_home_value, median_rent, white_population, black_population, hispanic_population, asian_population, foreign_born_population, health_insurance_coverage", "geography": "object (required) — Geographic scope for the query", "year": "integer — Survey year (default: 2023)", "dataset": "acs5|acs1 — ACS dataset: acs5 (5-year estimates, more areas) or acs1 (1-year, larger areas only). Default: acs5"}
- Output: {"year": "integer", "dataset": "string", "geography": "object", "data": "object — Variable values keyed by human-readable name"}

### Craigslist Search
- Description: Search Craigslist listings by city and category. Returns structured results with titles, prices, URLs, locations, dates, and thumbnails.
- Path: /api/tools/search-craigslist
- Price: 50 credits ($0.01)
- Input: {"query": "string (required) — Search query (e.g., 'bicycle', 'studio apartment')", "city": "string (required) — Craigslist city subdomain (e.g., 'sfbay', 'newyork', 'chicago', 'losangeles')", "category": "sss|apa|jjj|ggg|ccc|rrr — Listing category: sss=for sale, apa=apartments, jjj=jobs, ggg=gigs, ccc=community, rrr=rooms (default: sss)", "max_results": "integer — Maximum listings to return (default: 25)"}
- Output: {"listings": "array", "query": "string", "city": "string", "category": "string", "total_results": "integer", "search_url": "string"}

### Currency Rate
- Description: Convert between currencies using live exchange rates. Supports 30+ currencies including USD, EUR, GBP, JPY. Free, no rate limits.
- Path: /api/tools/currency-rate
- Price: 25 credits ($0.00)
- Input: {"from": "string (required) — Source currency code (e.g., 'USD', 'EUR', 'GBP')", "to": "string (required) — Target currency code (e.g., 'EUR', 'JPY', 'GBP')", "amount": "number — Amount to convert (default: 1)"}
- Output: {"from": "string", "to": "string", "amount": "number", "rate": "number — Exchange rate (1 unit of source currency)", "result": "number — Converted amount", "date": "string — Rate date (YYYY-MM-DD)"}

### Email - Check Inbox
- Description: Check your email inbox for replies and incoming messages. Requires a claimed email address via claim-email-address.
- Path: /api/tools/check-inbox
- Price: 25 credits ($0.00)
- Input: {"limit": "integer — Number of messages to return (default: 20, max: 50)"}
- Output: {"address": "string — Your claimed email address", "messages": "array"}

### Email - Claim Address
- Description: Claim a custom email address (@mail.agentpatch.ai) for your agent. Once claimed, you can send emails from this address and receive replies into your inbox.
- Path: /api/tools/claim-email-address
- Price: 5000 credits ($0.50)
- Input: {"handle": "string (required) — Email handle to claim (e.g. 'my-agent' → my-agent@mail.agentpatch.ai). Lowercase alphanumeric + hyphens, 3-30 chars."}
- Output: {"address": "string — Full email address (handle@mail.agentpatch.ai)", "status": "claimed"}

### Email - Send
- Description: Send transactional emails. Supports plain text and HTML content with custom sender names and reply-to addresses.
- Path: /api/tools/send-email
- Price: 100 credits ($0.01)
- Input: {"to": "string (required) — Recipient email address", "subject": "string (required) — Email subject line", "body": "string (required) — Plain text email body", "html": "string — Optional HTML email body", "from_name": "string — Optional sender display name", "reply_to": "string — Optional reply-to address"}
- Output: {"messageId": "string — Resend message ID", "status": "sent"}

### Generate Image (FLUX 2 Pro)
- Description: Generate high-quality images using FLUX 2 Pro. Premium model with excellent prompt adherence, text rendering, and photorealistic output.
- Path: /api/tools/generate-image-flux-2-pro
- Price: 600 credits ($0.06)
- Input: {"prompt": "string (required) — Text description of the image to generate", "image_size": "square_hd|square|portrait_4_3|portrait_16_9|landscape_4_3|landscape_16_9 — Image size preset (default: square_hd)", "guidance_scale": "number — How closely to follow the prompt (default: 3.5). Higher = more literal.", "output_format": "jpeg|png — Image format (default: jpeg)", "safety_tolerance": "integer — Content filter strictness, 1=strictest (default: 2)"}
- Output: {"image_url": "string — URL of the generated image", "content_type": "string — MIME type of the image", "width": "integer — Image width in pixels", "height": "integer — Image height in pixels"}

### Generate Image (FLUX Schnell)
- Description: Generate images using FLUX.1 Schnell. Extremely fast and cheap open-source model, great for quick drafts and iteration.
- Path: /api/tools/generate-image-flux-schnell
- Price: 60 credits ($0.01)
- Input: {"prompt": "string (required) — Text description of the image to generate", "image_size": "square_hd|square|portrait_4_3|portrait_16_9|landscape_4_3|landscape_16_9 — Image size preset (default: square_hd)", "num_inference_steps": "integer — Number of inference steps, 1-12 (default: 4). More steps = higher quality but slower.", "output_format": "jpeg|png — Image format (default: jpeg)"}
- Output: {"image_url": "string — URL of the generated image", "content_type": "string — MIME type of the image", "width": "integer — Image width in pixels", "height": "integer — Image height in pixels"}

### Generate Image (Recraft)
- Description: Generate high-quality images using Recraft v3. Supports realistic photos, digital illustrations, and vector art with configurable aspect ratios.
- Path: /api/tools/generate-image-recraft
- Price: 800 credits ($0.08)
- Input: {"prompt": "string (required) — Text description of the image to generate", "style": "realistic_image|digital_illustration — Image style (default: realistic_image)", "size": "square_hd|square|portrait_4_3|portrait_16_9|landscape_4_3|landscape_16_9 — Image size preset (default: square_hd)"}
- Output: {"image_url": "string — URL of the generated image", "content_type": "string — MIME type of the image", "file_size": "integer — File size in bytes"}

### Generate Image 4K (Nano Banana 2)
- Description: Generate 4K images using Google's Gemini Flash Image model via Nano Banana 2. Supports text rendering and multiple aspect ratios.
- Path: /api/tools/generate-image-nano-banana
- Price: 3200 credits ($0.32)
- Input: {"prompt": "string (required) — Text description of the image to generate", "aspect_ratio": "1:1|16:9|9:16|4:3|3:4|3:2|2:3|21:9 — Aspect ratio (default: 1:1)", "output_format": "png|jpeg|webp — Image format (default: png)"}
- Output: {"image_url": "string — URL of the generated image", "content_type": "string — MIME type of the image", "file_size": "integer — File size in bytes", "width": "integer — Image width in pixels", "height": "integer — Image height in pixels"}

### GitHub Repo Search
- Description: Search GitHub repositories by keyword. Filter by language and minimum stars, sort by stars, forks, or recency. Returns repo details including stars, forks, language, topics, and license. Great for finding open-source projects, competitive analysis, and technology landscape research.
- Path: /api/tools/github-repo-search
- Price: 50 credits ($0.01)
- Input: {"query": "string (required) — Search query (e.g. 'voice AI agent', 'markdown parser')", "language": "string — Filter by programming language (e.g. 'python', 'typescript')", "min_stars": "integer — Minimum star count (e.g. 100)", "sort": "stars|forks|updated|best-match — Sort order (default: best-match)", "limit": "integer — Number of results (default: 10, max: 30)"}
- Output: {"repos": "array", "total_count": "integer — Total matching repos", "query": "string — The search query used (including qualifiers)"}

### GitHub Repo Stats
- Description: Get statistics for any public GitHub repository. Returns stars, forks, open issues, language, topics, recent activity, and more. Useful for gauging open-source project traction and adoption.
- Path: /api/tools/github-repo-stats
- Price: 50 credits ($0.01)
- Input: {"owner": "string (required) — GitHub username or organization (e.g. 'facebook')", "repo": "string (required) — Repository name (e.g. 'react')"}
- Output: {"full_name": "string — Full repo name (owner/repo)", "description": "string — Repository description", "html_url": "string — GitHub URL", "stars": "integer — Star count", "forks": "integer — Fork count", "open_issues": "integer — Open issue count", "watchers": "integer — Watcher/subscriber count", "language": "string — Primary language", "license": "string — SPDX license ID", "created_at": "string — ISO 8601 creation date", "updated_at": "string — ISO 8601 last update", "pushed_at": "string — ISO 8601 last push", "size": "integer — Repo size in KB", "default_branch": "string", "is_fork": "boolean", "is_archived": "boolean", "topics": "array", "homepage": "string — Project homepage URL"}

### Google Image Search
- Description: Search Google Images and get structured results. Returns image URLs and source pages with optional filters for size, color, and type.
- Path: /api/tools/google-image-search
- Price: 75 credits ($0.01)
- Input: {"query": "string (required) — Image search query", "country": "string — Country code (e.g., 'us', 'uk', 'de')", "language": "string — Language code (e.g., 'en', 'es', 'fr')", "page": "integer — Page number (0-indexed)", "imgsz": "l|m|i|qsvga|vga|svga|xga — Image size filter (l=large, m=medium, i=icon)", "image_color": "bw|trans|red|orange|yellow|green|teal|blue|purple|pink|white|gray|black|brown — Color filter", "image_type": "face|photo|clipart|lineart|animated — Image type filter"}
- Output: {"results": "array", "query": "string"}

### Google Maps
- Description: Search Google Maps for businesses and places. Returns structured data including ratings, reviews, addresses, and contact info.
- Path: /api/tools/google-maps
- Price: 75 credits ($0.01)
- Input: {"query": "string (required) — Search query (e.g., 'coffee shops in NYC')", "coordinates": "string — Center coordinates (e.g., '@40.7,-74.0')", "zoom": "integer — Map zoom level", "language": "string — Language code (e.g., 'en')"}
- Output: {"results": "array", "query": "string"}

### Google News
- Description: Search Google News for recent articles and headlines. Returns structured news results with titles, snippets, sources, and publication times.
- Path: /api/tools/google-news
- Price: 75 credits ($0.01)
- Input: {"query": "string (required) — News search query", "country": "string — Country code (e.g., 'us', 'uk', 'de')", "language": "string — Language code (e.g., 'en', 'es', 'fr')", "results": "integer — Number of results (default: 10)", "page": "integer — Page number (0-indexed)"}
- Output: {"results": "array", "query": "string"}

### Google Reviews
- Description: Get Google reviews for any business or place. Returns individual reviews with ratings, dates, and text, plus overall place info. Use google-maps first to get the data_id, then pass it here. Great for competitive analysis, reputation research, and understanding customer sentiment.
- Path: /api/tools/google-reviews
- Price: 75 credits ($0.01)
- Input: {"data_id": "string (required) — Google Maps data ID (get this from the google-maps tool's data_id field)", "sort": "qualityScore|newestFirst|ratingHigh|ratingLow — Sort order for reviews (default: qualityScore)", "language": "string — Language code (e.g., 'en')"}
- Output: {"place": "object", "reviews": "array", "data_id": "string — The place ID used"}

### Google Search
- Description: Search Google and get structured results. Returns organic search results with titles, links, and snippets.
- Path: /api/tools/google-search
- Price: 50 credits ($0.01)
- Input: {"query": "string (required) — Search query", "country": "string — Country code (e.g., 'us', 'uk', 'de')", "language": "string — Language code (e.g., 'en', 'es', 'fr')", "results": "integer — Number of results (default: 10)", "page": "integer — Page number (0-indexed)"}
- Output: {"results": "array", "total_results": "integer", "query": "string"}

### Google Suggest
- Description: Get Google autocomplete suggestions for any query. Returns the search suggestions that appear when typing in Google.
- Path: /api/tools/google-suggest
- Price: 25 credits ($0.00)
- Input: {"query": "string (required) — Search query to get suggestions for", "country": "string — Two-letter country code (e.g. us, gb, de). Default: us", "language": "string — Language code (e.g. en, es, fr). Default: en"}
- Output: {"suggestions": "array — Autocomplete suggestions", "query": "string"}

### Google Trends
- Description: Get Google Trends data for search terms. Returns per-keyword summaries (baseline, peak, current, growth multiplier, trend direction) and a compact timeline. Compare up to 5 queries.
- Path: /api/tools/google-trends
- Price: 75 credits ($0.01)
- Input: {"query": "string (required) — Search term(s), comma-separated for comparison (max 5)", "data_type": "TIMESERIES|GEO_MAP|GEO_MAP_0 — Data type: interest over time, regional comparison, or regional interest (default: TIMESERIES)", "geo": "string — Location filter (e.g., 'US', 'GB', 'US-NY')", "region": "COUNTRY|REGION|DMA|CITY — Regional granularity for geo breakdowns", "language": "string — Language code (e.g., 'en', 'es')", "date": "string — Time range (e.g., 'now 1-d', 'today 12-m', 'all', or 'YYYY-MM-DD YYYY-MM-DD')", "cat": "integer — Category ID (0 = all categories)", "gprop": "images|news|froogle|youtube — Filter by property (default: web search)", "tz": "integer — Timezone offset in minutes (default: 420)"}
- Output: {"query": "string — The original query string", "keywords": "array — Per-keyword summary with growth metrics", "timeline": "array — Compact time series with one object per date, keys are keyword names with integer values (0-100)"}

### Hacker News Search
- Description: Search Hacker News stories and comments. Find discussions, product launches, technical posts, and community opinions by keyword. Defaults to stories only for higher relevance. Filter by content type, time range, minimum points, and sort by relevance or date.
- Path: /api/tools/hackernews-search
- Price: 50 credits ($0.01)
- Input: {"query": "string (required) — Search query text", "type": "story|comment|all — Filter by content type (default: story)", "sort": "relevance|date — Sort order (default: relevance)", "time_range": "last_24h|past_week|past_month|past_year — Filter to recent content (default: all time)", "limit": "integer — Number of results to return (default: 20, max: 50)", "min_points": "integer — Minimum points/upvotes to filter results (e.g. 5 to exclude low-quality posts)"}
- Output: {"hits": "array", "total_hits": "integer — Total matching results", "query": "string — The search query used"}

### Hello World
- Description: Free test tool to verify your AgentPatch setup. Returns a greeting message. No external API calls, always succeeds.
- Path: /api/tools/hello-world
- Price: 0 credits ($0.00)
- Input: {"name": "string — Name to greet (default: World)"}
- Output: {"message": "string — Greeting message", "timestamp": "string — ISO 8601 timestamp"}

### LinkedIn Jobs Search
- Description: Search LinkedIn job listings by job title or company name. Filter by location, job type, experience level, work model, and posting date.
- Path: /api/tools/linkedin-jobs
- Price: 75 credits ($0.01)
- Input: {"field": "string (required) — Job title or company name to search for (e.g. 'Product Manager' or 'Amazon')", "location": "string — Location to search in (e.g. 'San Francisco')", "geoid": "string — LinkedIn location ID. Default 92000000 (global)", "page": "integer — Page number (starts at 1)", "sort_by": "day|week|month — Filter by posting date: day (last 24h), week (last 7 days), month (last 30 days)", "job_type": "temporary|contract|volunteer|full_time|part_time — Type of employment", "exp_level": "internship|entry_level|associate|mid_senior_level|director — Required experience level", "work_type": "at_work|remote|hybrid — Work model"}
- Output: {"jobs": "array", "field": "string", "location": "string"}

### LinkedIn Profile
- Description: Scrape LinkedIn profile data. Returns structured profile information including experience, education, and about section.
- Path: /api/tools/linkedin-profile
- Price: 1500 credits ($0.15)
- Input: {"url": "string (required) — LinkedIn profile URL"}
- Output: {"profile": "object", "url": "string"}

### PDF to Text
- Description: Extract text from any PDF document by URL. Supports multi-page documents up to 50MB. Returns plain text content and page count.
- Path: /api/tools/pdf-to-text
- Price: 50 credits ($0.01)
- Input: {"url": "string (required) — URL of the PDF file to extract text from"}
- Output: {"text": "string — Extracted text content", "pages": "integer — Number of pages in the PDF", "url": "string — Source URL"}

### Paste
- Description: Upload text, HTML, or JSON and get a public URL that auto-expires. Perfect for sharing logs, reports, previews, or any small text artifact. Content is hosted for up to 7 days.
- Path: /api/tools/paste
- Price: 25 credits ($0.00)
- Input: {"content": "string (required) — Text content to host (max 1MB)", "content_type": "text/plain|text/html|application/json — MIME type of the content (default: text/plain)", "ttl_hours": "integer — Hours until expiry (default: 72, max: 168 = 7 days)"}
- Output: {"url": "string — Public URL to view the paste", "id": "string — Paste ID", "content_type": "string — MIME type of the hosted content", "expires_at": "string — ISO 8601 expiry timestamp"}

### Product Hunt Search
- Description: Search Product Hunt by keyword or browse launches by topic. Use query for keyword search or topic for browsing a specific category. Returns products with titles, descriptions, and links. Useful for discovering competitors, validating product ideas, and gauging launch traction. Popular topics: artificial-intelligence, developer-tools, productivity, marketing, analytics, design-tools, saas, open-source, api-1, no-code, fintech, health-fitness, education, chrome-extensions, github, machine-learning, chatgpt, automation, email, social-media-tools.
- Path: /api/tools/producthunt-search
- Price: 50 credits ($0.01)
- Input: {"query": "string — Keyword search query (e.g. 'AI phone assistant'). Provide query or topic.", "topic": "string — Topic slug to browse (e.g. 'artificial-intelligence', 'developer-tools'). Use kebab-case.", "limit": "integer — Number of results to return (default: 10, max: 20)", "order": "VOTES|NEWEST|RANKING — Sort order (default: RANKING)", "featured": "boolean — Filter to only featured (front-page) products", "posted_after": "string — ISO 8601 date to filter products posted after (e.g. '2026-01-01')", "posted_before": "string — ISO 8601 date to filter products posted before"}
- Output: {"results": "array — Keyword search results (when query is used)", "posts": "array — Full product details (when topic is used)", "query": "string — The keyword search query used", "topic": "string — The topic slug browsed"}

### PyPI Package Stats
- Description: Get metadata and download statistics for any Python package on PyPI. Returns version info, daily/weekly/monthly downloads, dependencies, and more. Useful for measuring Python library adoption.
- Path: /api/tools/pypi-package-stats
- Price: 50 credits ($0.01)
- Input: {"package": "string (required) — PyPI package name (e.g. 'requests' or 'anthropic')"}
- Output: {"name": "string", "description": "string", "latest_version": "string", "homepage": "string", "repository_url": "string", "license": "string", "author": "string", "requires_python": "string", "downloads_last_day": "integer", "downloads_last_week": "integer", "downloads_last_month": "integer", "keywords": "array", "classifiers": "array", "last_published": "string — ISO 8601 timestamp of latest version"}

### Reddit Search
- Description: Search Reddit posts and comments. Returns structured results with scores, comments, and post content.
- Path: /api/tools/reddit-search
- Price: 300 credits ($0.03)
- Input: {"query": "string (required) — Search query", "subreddit": "string — Limit search to specific subreddit", "sort": "relevance|hot|top|new|comments — Sort order (default: relevance)", "time": "hour|day|week|month|year|all — Time filter (default: all)", "limit": "integer — Number of results (default: 25)"}
- Output: {"posts": "array", "query": "string", "subreddit": "string"}

### SEC Company Financials
- Description: Get financial data for a public company from SEC EDGAR filings (10-K and 10-Q). Use sec-company-search to find CIK numbers. Common XBRL fact names: Revenues, RevenueFromContractWithCustomerExcludingAssessedTax, NetIncomeLoss, GrossProfit, OperatingIncomeLoss, Assets, Liabilities, StockholdersEquity, CashAndCashEquivalentsAtCarryingValue, LongTermDebt, EarningsPerShareBasic, EarningsPerShareDiluted, CommonStockSharesOutstanding, OperatingExpenses, ResearchAndDevelopmentExpense, CostOfGoodsAndServicesSold, InterestExpense, IncomeTaxExpenseBenefit, AccountsReceivableNetCurrent, InventoryNet, PropertyPlantAndEquipmentNet, Goodwill, RetainedEarningsAccumulatedDeficit, DividendsCommonStockCash, PaymentsOfDividendsCommonStock, ShareBasedCompensation. If omitted, defaults to: Revenues, NetIncomeLoss, Assets, Liabilities, StockholdersEquity, EarningsPerShareBasic, OperatingIncomeLoss, CashAndCashEquivalentsAtCarryingValue.
- Path: /api/tools/sec-company-financials
- Price: 75 credits ($0.01)
- Input: {"cik": "string (required) — SEC CIK number (e.g., '0000320193' for Apple)", "facts": "array — XBRL fact names to return (e.g., 'Revenues', 'NetIncomeLoss', 'Assets'). Defaults to common financial metrics if omitted.", "years": "integer — Years of data to return (default: 5)"}
- Output: {"company": "object", "facts": "object — Financial facts keyed by XBRL tag name, each containing an array of period entries"}

### SEC Company Search
- Description: Search for public companies in the SEC EDGAR database. Find company CIK numbers, tickers, and names. Use the CIK with sec-company-financials to get financial data.
- Path: /api/tools/sec-company-search
- Price: 50 credits ($0.01)
- Input: {"query": "string (required) — Company name or ticker symbol (e.g., 'Apple', 'AAPL')"}
- Output: {"results": "array"}

### Scrape Web
- Description: Scrape any webpage and get clean markdown content. Strips scripts, styles, navigation, and other non-content elements. Supports JavaScript rendering and premium proxies for difficult sites.
- Path: /api/tools/scrape-web
- Price: 200 credits ($0.02)
- Input: {"url": "string (required) — URL to scrape", "render_js": "boolean — Enable JavaScript rendering (default: false)", "premium": "boolean — Use premium residential proxies (default: false)", "country": "string — Country code for geo-targeting (e.g., 'us', 'uk')"}
- Output: {"markdown": "string — Page content converted to clean markdown", "url": "string — URL that was scraped"}

### Stock Quote
- Description: Get real-time stock quotes. Returns current price, change, and daily high/low for US stocks.
- Path: /api/tools/stock-quote
- Price: 200 credits ($0.02)
- Input: {"symbol": "string (required) — Stock ticker symbol (e.g., AAPL, MSFT)"}
- Output: {"symbol": "string", "current_price": "number", "change": "number", "percent_change": "number", "high": "number", "low": "number", "open": "number", "previous_close": "number", "timestamp": "number"}

### US Trade Data
- Description: Get US international trade (import/export) data from the Census Bureau. Query by Harmonized System commodity code, country, and time period. Returns monthly and year-to-date trade values in USD. Common country codes: 5700=China, 2010=Canada, 2900=Mexico, 4759=Germany, 5880=Japan, 4120=United Kingdom, 6091=South Korea, 3510=India, 1220=Australia. Commodity levels: HS2 (broad categories like '85'=Electrical Machinery), HS4 (subcategories like '8517'=Telephones), HS6 (specific products like '851712'=Cell Phones).
- Path: /api/tools/trade-data
- Price: 75 credits ($0.01)
- Input: {"direction": "imports|exports (required) — Trade direction: 'imports' or 'exports'", "commodity_level": "HS2|HS4|HS6 (required) — Harmonized System aggregation level. HS2=broad (99 categories), HS4=subcategories, HS6=specific products", "time": "string (required) — Month to query in YYYY-MM format (e.g., '2024-06'). Data available from 2013-01 onwards.", "country_code": "string — 4-digit Census country code to filter by (e.g., '5700' for China). Omit for all countries.", "commodity_code": "string — HS commodity code to filter by (e.g., '85' for Electrical Machinery at HS2, '8517' for Telephones at HS4). Omit for all commodities."}
- Output: {"direction": "string", "commodity_level": "string", "time": "string", "country_code": "string", "total_rows": "integer — Total rows from the API", "rows_returned": "integer — Rows returned (capped at 50)", "rows": "array"}

### Weather
- Description: Get current weather for any city worldwide. Returns temperature, humidity, wind speed, and conditions with metric or imperial units.
- Path: /api/tools/weather
- Price: 25 credits ($0.00)
- Input: {"location": "string (required) — City name (e.g., 'London', 'New York', 'Tokyo')", "units": "metric|imperial|standard — Temperature units: metric (Celsius), imperial (Fahrenheit), standard (Kelvin). Default: metric"}
- Output: {"location": "string", "country": "string", "temperature": "number", "feels_like": "number", "humidity": "number", "description": "string", "wind_speed": "number", "units": "string", "coordinates": "object"}

### YouTube Transcript
- Description: Fetch the transcript of a YouTube video. Returns timestamped text segments with start times and durations.
- Path: /api/tools/youtube-transcript
- Price: 50 credits ($0.01)
- Input: {"video_id": "string (required) — YouTube video ID (e.g., 'dQw4w9WgXcQ')", "language": "string — Language code for transcript (e.g., 'en', 'es')"}
- Output: {"video_id": "string", "segments": "array"}

### arXiv Search
- Description: Search arXiv for academic papers and preprints. Find research papers by keyword, filter by subject category, and sort by relevance or date. Covers CS, AI/ML, physics, math, biology, economics, and more. Common categories: cs.AI (Artificial Intelligence), cs.LG (Machine Learning), cs.CL (Computation & Language/NLP), cs.CV (Computer Vision), cs.CR (Cryptography & Security), cs.SE (Software Engineering), stat.ML (Statistics - ML), econ.GN (General Economics), q-bio (Quantitative Biology).
- Path: /api/tools/arxiv-search
- Price: 50 credits ($0.01)
- Input: {"query": "string (required) — Search query (e.g. 'large language models', 'reinforcement learning from human feedback')", "category": "string — arXiv category to filter by (e.g. 'cs.AI', 'cs.LG', 'stat.ML')", "sort_by": "relevance|lastUpdatedDate|submittedDate — Sort order (default: relevance)", "sort_order": "ascending|descending — Sort direction (default: descending)", "limit": "integer — Number of results (default: 10, max: 50)"}
- Output: {"papers": "array", "total_results": "integer — Total matching papers", "query": "string — The search query used", "category": "string — Category filter applied (if any)"}

### eBay Product
- Description: Scrape an eBay product page. Returns structured product data including price, condition, seller info, shipping, returns, and specifications.
- Path: /api/tools/ebay-product
- Price: 400 credits ($0.04)
- Input: {"url": "string (required) — eBay product page URL"}
- Output: {"title": "string", "itemId": "string", "price": "string", "extracted_price": "number", "condition": "string", "condition_snippet": "string", "is_buy_it_now": "boolean", "seller": "object", "shipping_details": "object", "return_details": "object", "specifications": "array", "rating": "integer", "reviews": "integer", "url": "string"}

### eBay Search
- Description: Search eBay listings by keyword. Returns structured results with prices, seller info, condition, and shipping details.
- Path: /api/tools/ebay-search
- Price: 400 credits ($0.04)
- Input: {"query": "string (required) — Search query (e.g. 'iphone 15 pro')", "page": "integer — Page number (default: 1)"}
- Output: {"results": "array", "query": "string"}

### npm Package Stats
- Description: Get metadata and download statistics for any npm package. Returns version info, weekly/monthly downloads, maintainers, and more. Useful for measuring JavaScript/TypeScript library adoption.
- Path: /api/tools/npm-package-stats
- Price: 50 credits ($0.01)
- Input: {"package": "string (required) — npm package name (e.g. 'express' or '@anthropic-ai/sdk')"}
- Output: {"name": "string", "description": "string", "latest_version": "string", "homepage": "string", "repository_url": "string", "license": "string", "weekly_downloads": "integer", "monthly_downloads": "integer", "last_published": "string — ISO 8601 timestamp of latest version publish", "created_at": "string — ISO 8601 timestamp of first publish", "keywords": "array", "maintainers": "array"}

## Skill (Recommended)

A skill teaches your agent how and when to use AgentPatch, including prompt context for discovery and CLI usage.

### Claude Code
/plugin marketplace add fullthom/agentpatch-claude-skill
/plugin install agentpatch@agentpatch

### OpenClaw
clawhub install agentpatch

Works with any agent that supports skills.

## CLI

The AgentPatch CLI is designed for AI agents to use via shell access. Install it, and your agent can discover and invoke any tool on the marketplace.

pip install agentpatch

ap search "image generation"
ap info google-search
ap run google-search --input '{"query": "test"}'
ap job job_abc123 --poll

Set AGENTPATCH_API_KEY env var, or run: ap config set-key
Full docs: https://github.com/fullthom/agentpatch-python

## MCP Setup (Alternative)

Alternatively, connect via MCP (if your agent doesn't support skills or shell access):

### OpenClaw
Install skill (recommended): clawhub install agentpatch
Or add to ~/.openclaw/openclaw.json:
{
  "mcp": {
    "servers": {
      "agentpatch": {
        "transport": "streamable-http",
        "url": "https://agentpatch.ai/mcp"
      }
    }
  }
}

### Claude Code
Install the AgentPatch skill (recommended):
/plugin marketplace add fullthom/agentpatch-claude-skill
/plugin install agentpatch@agentpatch

Or add as MCP server:
claude mcp add -s user --transport http agentpatch https://agentpatch.ai/mcp \
  --header "Authorization: Bearer YOUR_API_KEY"

### Cursor
Open Settings > MCP > Add new global MCP server.
Add to ~/.cursor/mcp.json:
{
  "mcpServers": {
    "agentpatch": {
      "url": "https://agentpatch.ai/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

### Codex
Add to ~/.codex/config.toml:
[mcp_servers.agentpatch]
url = "https://agentpatch.ai/mcp"
bearer_token_env_var = "AGENTPATCH_API_KEY"

Then: export AGENTPATCH_API_KEY=your_api_key

### Generic MCP Client
Connect to the MCP endpoint:
- Streamable HTTP: https://agentpatch.ai/mcp
- SSE: https://agentpatch.ai/mcp/sse

## REST API

1. List all tools (no auth required):
curl "https://agentpatch.ai/api/tools"

2. Search for tools (no auth required):
curl "https://agentpatch.ai/api/search?q=email"

3. Have your user sign up at https://agentpatch.ai/login and create an API key.
After signing up, top up credits to start using tools.

4. Invoke a tool:
curl -X POST "https://agentpatch.ai/api/tools/{slug}" \
  -H "Authorization: Bearer $AGENTPATCH_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "example"}'

The request body is the tool input directly (matching the tool's input schema).

5. (If async) Poll for result:
curl -H "Authorization: Bearer $AGENTPATCH_API_KEY" "https://agentpatch.ai/api/jobs/{job_id}"

## Auth

Sign up at https://agentpatch.ai/login
Go to Dashboard > API Keys to create an API key.
Pass it as: Authorization: Bearer your_api_key

## Credits

1 credit = $0.0001 USD (10,000 credits = $1.00)
Top up at Dashboard > Billing, or via API: POST /api/my/topup
Minimum $10, maximum $500.
Failed invocations (5xx, timeout) are fully refunded.

## API Reference

GET /api/tools — List all tools (no auth)
GET /api/search?q={query} — Search tools (no auth)
GET /api/tools/{slug} — Get tool details (no auth)
POST /api/tools/{slug} — Invoke a tool (auth required)
GET /api/jobs/{job_id} — Get job status and output (auth required)

Full OpenAPI spec: https://agentpatch.ai/api/doc
Swagger UI: https://agentpatch.ai/api/docs