Browse 160,118 AI agents

By Ali Abdullah

Quick overview This workflow scans a Google Drive folder of contract PDFs on a weekly schedule, extracts key renewal terms from new documents using the Gemini API, stores and updates a contract registry in Google Sheets, and sends renewal alerts via Gmail and Slack plus a weekly digest email. How it works Runs weekly (or manually for testing) and loads configuration values like Drive folder ID, Google Sheet ID, alert email, and Slack channel. Searches Google Drive for PDFs in the target folder and reads the current contract registry from Google Sheets. Compares Drive files to the registry to identify new contracts, then downloads each new PDF, extracts its text, and calls the Gemini API to extract structured contract terms. Appends extracted contract records to the Google Sheets Registry and combines new and existing contracts into a single list. Calculates days-to-renewal and assigns urgency tiers (90/60/30/notice_passed) with risk flags for short-notice auto-renewals, then filters contracts that have entered a new alert tier since the last run. For each alert due, builds a one-page renewal briefing, emails it via Gmail, posts a Slack notification, and updates the registry to record the last alerted tier. Builds a portfolio summary of tracked contracts, new contracts, alerts sent, and items needing review, emails the weekly digest via Gmail, and logs the run to a RunLog sheet in Google Sheets. Setup Create a Google Sheet with a Registry tab containing columns: file_id, vendor_name, contract_title, start_date, end_date, auto_renewal, notice_period_days, annual_value, currency, termination_penalty, key_contact, last_alert_tier, status, added_date, and a RunLog tab with date, contracts_scanned, new_contracts, alerts_sent. Add Google Drive, Google Sheets, Gmail, and Slack credentials in n8n, and set a Slack channel to receive renewal notifications. Get a Gemini API key (Google AI Studio) and fill in the Workflow Configuration values for geminiApiKey, driveFolderId, registrySheetId, alertEmail, slackChannelId, and maxCharsPerContract. Put your contract PDFs into the configured Google Drive folder and run the workflow once manually to populate the registry and validate alerts and digest emails. An n8n automation workflow template by Ali Abdullah.

By EcomOpsKits

Quick overview Compare price and availability changes for variants of one known public Shopify product. Save the latest complete report in an n8n Data Table and use it on the next manual run. This free workflow uses a paid, publisher-owned RapidAPI service. How it works A manual execution validates the product, currency, date and explicit spending consent. It reads the latest complete report from a Data Table, reserves one attempt and the request's recovery key, and verifies that reservation before calling the API once. It validates the full response, saves it together with the updated pilot state and reads it back before returning success. The first success establishes a baseline; later successes include variant-level changes. Errors preserve the prior report. No scheduled monitoring or alerts are included. Only one execution and one workflow copy may use the history row. Setup Use n8n2.40.7 or a compatible release supporting Data Table create/get/upsert. Enter a known public product URL, its expected currency and a stop date within 14 days. Add your RapidAPI key as an HTTP Header Auth credential named x-rapidapi-key and select it on the HTTP node. Review current pricing and the complete setup note before enabling spending consent. Run one execution at a time, one workflow copy, without manual history edits. Stop at 14 reserved attempts or the saved date; follow the note's exact recovery procedure. Additional info EcomOpsKits publishes the linked paid API: https://rapidapi.com/ecomclosekit/api/shopify-variant-price-and-availability-diff The USD0.01/request tariff is the October4 publication snapshot, not a current buyer quote; bandwidth/other charges and n8n hosting are separate. Replays/errors may bill. The workflow counter is not a billing cap or concurrency lock. Full product URL/previous report data is transmitted to RapidAPI/Cloudflare/Apify; no retention period is promised. Only one public product/source/currency and 1–249variants are supported. This is a latest-report comparison pilot, without sales/inventory inference, currency conversion, schedules or alerts. Free direct Shopify JSON/code solutions are possible; this API is a convenience choice. On interruption/error before commit, inspect the saved row and prior execution. If recovery is appropriate, enable recoverPendingRequest and run the full workflow with the same credential/consumer. It reuses the exact saved body/key, consumes another slot and may bill. A cached provider error remains an error; recovery does not promise a fresh observation. An uncertain backend start may need provider reconciliation. Never use native Retry execution, partial Execute step on HTTP/save nodes, a fresh key or reset history to recover: these can bypass the reservation or create another billable request. After a successful commit, pending is already cleared; explicit recovery then refuses another call. Disable recovery only when deliberately starting a new observation. An n8n automation workflow template by EcomOpsKits.

By CleanScrape

Quick overview This workflow runs every Monday, scrapes the last 7 days of 1–2 star reviews from the App Store and Google Play via Apify, summarizes the main complaint themes with OpenAI, and posts a weekly digest to a Slack channel only when complaints exist. How it works Runs every Monday at 9:00 on a schedule trigger. Sends your App Store and Google Play app identifiers, country, rating filter (1–2 stars), and a 7-day lookback window to the Apify “App Store & Google Play Reviews Scraper” actor and retrieves the resulting review dataset. Aggregates the reviews into a single weekly payload, counting items per store and formatting up to 150 review lines with store, rating, version, and text for summarization. Stops if there are no matching reviews for the week. Uses OpenAI (gpt-5-mini) to produce a short Slack-ready digest listing 3–5 top complaint themes with estimated counts, a representative quote, and version callouts where relevant. Posts the digest to the selected Slack channel with a header showing total review count and the App Store vs Google Play breakdown. Setup Add an Apify credential (API token) and ensure the workflow can run the “cleanscrape/app-store-reviews-scraper” actor. Add an OpenAI API credential for the chat model used to summarize the reviews. Add a Slack credential with permission to post messages and make sure the target channel exists. Update the Settings values for your Google Play and App Store app links/IDs, store country, how many reviews to check, and the Slack channel name. Requirements Apify account. The Actor charges per review saved (about $0.20 per 1,000 reviews plus a small start fee); Apify's free plan includes monthly credit. OpenAI API key. Slack workspace and a channel for the digest. Customization Change the star ratings or the 7-day window in the Apify node's input. Check more reviews per run with reviewsToCheck in Settings. Send the digest to email, Microsoft Teams or Notion instead of Slack. Additional info The example Settings use the Strava app; replace them with your own app's links. Schedule times use the workflow's time zone. Made by CleanScrape, the developer of the Apify Actor used here. An n8n automation workflow template by CleanScrape.

By Andy

Quick overview Create a quotation draft from a request CSV and an approved price list, with unmatched or conflicting items held for review. Built for parts distributors and sales teams using exact SKUs. How it works Receives an n8n Form submission with two uploaded CSV files (request and price list) and a chosen delimiter. Validates that both uploads are UTF-8 .csv files within size limits and reads their contents as text. Parses both CSVs, enforces required headers, and builds a catalogue from the price list while flagging invalid rows and blocking duplicate SKUs. Validates each request line (line_id, quantity, unit) and matches it to exactly one case-sensitive SKU in the catalogue, holding unknown, blocked, duplicate, or unit-mismatched lines for review. Calculates unit prices and line totals for accepted lines, keeps separate totals per currency, and generates report.json plus quotation-draft.csv and review.csv. Packages the three files into a single ZIP and returns it as the form completion download with a status summary. Setup Import the workflow and run it from the editor to open the n8n Form trigger. Prepare two UTF-8 CSV files matching the required schemas (request: line_id, sku, quantity, unit; price list: sku, description, unit, unit_price, currency). Upload both CSVs in the form and select the delimiter that both files share (comma, semicolon, or tab). Submit the form and download quotation-review.zip. Inspect report.json and review.csv, resolve all reported issues, and check the accepted line totals before using the draft. A partial draft is not a complete offer. Requirements Tested on n8n 2.28.6. Built-in nodes only; no API credentials or paid services. Existing hosting costs apply. Both files must be UTF-8 CSVs using the same delimiter. Limits: 4,000,000 bytes per file, 1,000 request rows, 10,000 catalogue rows. Supported currencies: USD, EUR, GBP, THB, JPY, KWD. Customization Change the input files and delimiter in the form. Rename source headers to the documented schema. Validation changes require code edits and retesting. Additional info CSV-only: no Google Sheets connection, PDF extraction, tax calculation, or customer delivery. Unknown, duplicate, conflicting, and invalid items remain unpriced for human review. A partial draft is not a complete offer. Enable form authentication before publishing a production form. An n8n automation workflow template by Andy.

By James Mundy

Quick overview This workflow takes a lead URL, uses Gaffa to extract contact details from the web page, and then creates or updates a HubSpot contact based on the extracted email. How it works Runs when you manually execute the workflow. Sets the lead URL and the CSS selector that defines which part of the page to parse. Builds a Gaffa request that defines a JSON schema for first name, last name, company, role, and email. Sends the request to Gaffa to parse the target web page and return the extracted contact fields. Upserts the contact in HubSpot using the extracted email and maps the other extracted fields to HubSpot contact properties. Setup Install the community node package @gaffa-dev/n8n-nodes-gaffa in your self-hosted n8n instance. Create and add a Gaffa API credential with your API key from gaffa.dev. Create a HubSpot private app with crm.objects.contacts read/write scopes and add its app token as HubSpot credentials. Update the lead URL (and optionally the selector) in the workflow settings before running it. An n8n automation workflow template by James Mundy.

By Jason Bullen

Quick overview This workflow runs a sample support-ticket triage in n8n and uses FirstHelm’s control plane APIs to log activity, evaluate communication constraints, and route high-urgency tickets for human approval while logging completion. How it works Starts when you manually execute the workflow in n8n. Creates a sample support ticket payload with customer, subject, urgency, and ticket ID. Sends a log event to FirstHelm (logActivity) to record that triage has started. Calls FirstHelm (evaluateConstraint) to check whether sending a resolution email to the customer is allowed under the configured constraints. If the ticket urgency is high, requests human approval via FirstHelm (requestApproval) and logs that the workflow is waiting for an operator decision. If the ticket urgency is not high, logs an auto-response email action to FirstHelm. Logs a final completion event to FirstHelm (logActivity) to mark the triage as finished. Setup Create a FirstHelm account and obtain a FirstHelm API key, then replace YOUR_FIRSTHELM_API_KEY in all HTTP requests. Set your FirstHelm agent_id by replacing YOUR_AGENT_ID in the request bodies. (Optional) Replace the sample ticket fields (customer, subject, urgency, ticket_id) with values from your real support system or upstream trigger. Requirements Free FirstHelm account (firsthelm.dev) + API key Customization Swap the sample ticket trigger for your own Gmail/Outlook/helpdesk node. An n8n automation workflow template by Jason Bullen.

By Siddharth Gupta

Quick Overview This workflow scrapes job listings from multiple company career sites, stores and updates them in Google Sheets, cleans and extracts job text from each job page, then uses Google Gemini and Groq (via LangChain) to generate LinkedIn-ready job posts that can be published through LinkedIn and reported via Telegram. How it works Triggers from a Telegram bot command or on a scheduled cron timetable. Scrapes job listings from multiple career sites (including TruGreen, SavATree, Perennial Services Group, Senske/UKG, Paylocity job boards, Breezy HR, and SilkRoad) and appends or updates new jobs in Google Sheets using Req ID as the key. Loads unprocessed jobs from Google Sheets, validates each record has an Original URL, and routes it to the correct brand-specific scraper based on the URL pattern. Fetches each job detail page, sanitizes the HTML into clean plain text, and writes the result back to Google Sheets with “Text created” and “Date Text Created” fields. For ad-hoc or scheduled posting, loads a job with cleaned text, validates the “Job Text” field is present, and prompts a LangChain agent using Google Gemini and Groq to draft a compliance-focused LinkedIn post. Publishes approved posts to a LinkedIn organization and logs the LinkedIn post URL, published status, post text, and publish date back into Google Sheets. Sends Telegram updates for process status and generates total, daily, and weekly posting reports from Google Sheets metrics. Setup Connect credentials for Google Sheets OAuth2, LinkedIn OAuth2 (with permission to post as your target organization), Telegram bot API, Groq API, and Google Gemini (PaLM) API. Create or select the Google Sheets spreadsheet and worksheet with the expected columns (for example: Req ID, Job title, Brand Name, City, State, Zip code, Address, Original URL, Job Text, Text created, Date Text Created, Published?, Post Link, Post text, Publish Date). Update the Google Sheets document ID, sheet/tab name, and any matching columns (Req ID and/or Original URL) to align with your spreadsheet. Set your Telegram chat IDs and configure the Telegram bot in BotFather, then start the bot and use the inline menu options to run scraping, cleaning, posting, or reporting. Confirm the LinkedIn organization ID used for posting matches your organization and that the workflow’s schedule times match your desired timezone and cadence. An n8n automation workflow template by Siddharth Gupta.

By starkAhmed43

An AI-powered application hosted on Hugging Face Spaces by starkAhmed43. Built with docker.

By cetiennec

Slides, exercise sheets and interactive spaces of my courses Hosted on Hugging Face Spaces by cetiennec. SDK: static.

By vllm-sr

Point at it, extract it, check it. A 0.25B encoder on CPU. Hosted on Hugging Face Spaces by vllm-sr. SDK: docker.

By mrfakename

Chat with Recursive's Flint coding model Hosted on Hugging Face Spaces by mrfakename. SDK: gradio.

By hugging-apps

Discover closed-form laws of nature from ≤10 data pairs Hosted on Hugging Face Spaces by hugging-apps. SDK: gradio.

By smirki

Expressive voice cloning that sighs, gasps and breaks Hosted on Hugging Face Spaces by smirki. SDK: gradio.

By samliumay

Inspect LLM-generated SQL before it reaches the database Hosted on Hugging Face Spaces by samliumay. SDK: docker.

By samliumay

A deliberately imperfect LLM bank assistant to pentest Hosted on Hugging Face Spaces by samliumay. SDK: docker.

By AlekseyCalvin

Run & Train Small LMs Easy & Free w/ GGUF, PEFT & more! Hosted on Hugging Face Spaces by AlekseyCalvin. SDK: gradio.

By jbhati305

An AI-powered application hosted on Hugging Face Spaces by jbhati305. Built with static.

By Roman81

An AI-powered application hosted on Hugging Face Spaces by Roman81. Built with static.

By SabaPivot

An AI-powered application hosted on Hugging Face Spaces by SabaPivot. Built with static.

By hugging-apps

Small cognition models predict your choices, trial by trial Hosted on Hugging Face Spaces by hugging-apps. SDK: gradio.

By xal2077

Video generation with a synchronized soundtrack Hosted on Hugging Face Spaces by xal2077. SDK: static.

By sotinhoss

Compare three reference-guided manga colorizers Hosted on Hugging Face Spaces by sotinhoss. SDK: gradio.

By mirrash7

An AI-powered application hosted on Hugging Face Spaces by mirrash7. Built with docker.

By mioulin

Adversarial evaluation toolkit for frontier language models Hosted on Hugging Face Spaces by mioulin. SDK: gradio.

By spark-nlp

An AI-powered application hosted on Hugging Face Spaces by spark-nlp. Built with docker.

By wb-droid

An AI-powered application hosted on Hugging Face Spaces by wb-droid. Built with gradio.

By GeorgyVlasov

An AI-powered application hosted on Hugging Face Spaces by GeorgyVlasov. Built with gradio.

By sms07

An AI-powered application hosted on Hugging Face Spaces by sms07. Built with streamlit.

By joelniklaus

Which coding-agent harness should you use for your model? Hosted on Hugging Face Spaces by joelniklaus. SDK: docker.

By cyun9286

An AI-powered application hosted on Hugging Face Spaces by cyun9286. Built with gradio.

By CharlesW

Confucius4-TTS with optional reference transcription Hosted on Hugging Face Spaces by CharlesW. SDK: gradio.

By hugging-apps

Step into a live, steerable generated world Hosted on Hugging Face Spaces by hugging-apps. SDK: gradio.

By vllm-sr

Towards Open Multimodal Foundation Decision Models Hosted on Hugging Face Spaces by vllm-sr. SDK: static.

By SalahALHaismawi

Arabic OCR with faithful tashkeel: #1 on KITAB & Misraj Hosted on Hugging Face Spaces by SalahALHaismawi. SDK: gradio.

By witness-archive

An AI-powered application hosted on Hugging Face Spaces by witness-archive. Built with static.

By hugging-apps

Image to high-res 3D mesh with Hi3Dv1 refinement Hosted on Hugging Face Spaces by hugging-apps. SDK: gradio.

By smolagents

An AI-powered application hosted on Hugging Face Spaces by smolagents. Built with docker.

By CodeSoft

Blind ranking of small language models! Hosted on Hugging Face Spaces by CodeSoft. SDK: gradio.

By za.co.theahg

Search and read AtoM archival catalogues with citable links; staff propose changes for review.

By watch.pcprice

Real eBay sold prices for used GPUs, CPUs, RAM and more in 7 markets, and eBay listing price checks.

By wiki.rust

Live data for the game Rust: skin prices and history, items, item store, loot tables, servers.

By wiki.wowdb

Search WoW Forever items, recipes, abilities and quests with page links from Azeroth DB.

By wine.nevini

Search the wines of NEVINI, a Dutch wine importer, and build Shopify carts with a checkout link.

By work.dutiful

Cited company background reports for agents from official registries. $3 per report via x402.

By work.timero

Time tracking for agencies: log time, run the timer, check client budgets and pull reports.

By xyz.dolphinamp

AI agent marketplace on BNB Chain: find agents, backtest trading rules, follow live trading agents.

By xyz.unfee

Flat 1-credit Base/ETH/XRPL RPC (eth_getLogs=1 credit), x402 wash filter, 0% checkout & tax CSV.

By ua.com.semeropost

Express delivery to/from Ukraine via DHL, FedEx & others: price estimates and customs guidance