Yearly Pay yearly and get 2 months free on every paid plan. See plans
Sign in Start free

AI workflows: Development

Ready-to-run AI workflows in the Development category: open one, fill in the short form and run it. Each one shows a sample of what it makes and what a run costs.

40 templates · 39 templates with a sample Clear filters
Sample outputText
Overall purpose This function accepts two values, a and b, and returns the result of applying JavaScript’s + operator to them. It can perform numeric addition or string concatenation, depending on the argument types. Section-by-section walkthrough function add(a, b) declares a function named add with two parameters: a and b. { ... } contains the function body. return sends the computed value back to the caller. a + b applies the + operator to the two arguments. The final ; terminates the return statement. For example: add(2, 3); // 5…
Read the full sample →

Code explainer

Input: function add(a, b) { return a + b; }

≈ 1,000
Use this
Sample outputText
Findings Missing input validation / type confusion Severity: Low Line ref: 1 Exploit sketch: Calling add("1", "2") returns "12" because + performs string concatenation when either operand is a string. Fix: Declare numeric parameter and return types, validate values at runtime, and reject non-finite numbers. Injection Severity: None identified Line ref: N/A Exploit sketch: The function does not pass input to an interpreter, database, shell, HTML sink, or other injection-sensitive context.…
Read the full sample →

Security review (lite)

Input: function add(a, b) { return a + b; }

≈ 1,000
Use this
Sample outputText
WITH customer_totals AS ( SELECT o.customer_id, SUM(o.total) AS total_order_value FROM orders AS o WHERE o.created_at = CURRENT_TIMESTAMP - INTERVAL '90' DAY GROUP BY o.customer_id ) SELECT c.id, c.name, ct.total_order_value FROM customer_totals AS ct JOIN customers AS c ON c.id = ct.customer_id ORDER BY ct.total_order_value DESC FETCH FIRST 10 ROWS ONLY; ```…
Read the full sample →

SQL query builder

Input: top 10 customers by total order value in the last 90 days, tables: customers(id,name), orders(id,customer_id,total,created_at)

≈ 1,000
Use this
Sample outputText
name: CI and Staging Deploy on: pull_request: push: branches: - dev Prevent older runs for the same branch or pull request from wasting resources. concurrency: group: ci-${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }} cancel-in-progress: true Use least-privilege defaults; individual jobs can add permissions if needed. permissions: contents: read env: NODE_VERSION: "22" PNPM_VERSION: "10" jobs: install: name: Install dependencies runs-on: ubuntu-latest steps: - name: Check out repository…
Read the full sample →

CI pipeline writer

Input: Node.js monorepo, pnpm, needs lint + test + build, deploy to staging on merge to dev

≈ 1,000
Use this
Sample outputText
Rate Limiting Design 1. Algorithm Choice Use a hierarchical token bucket as the primary enforcement algorithm, supplemented by a short sliding-window burst detector. Why Token Bucket Token bucket is the best primary mechanism because it: Allows legitimate short bursts while enforcing a sustainable average rate. Requires constant memory and near-constant processing per bucket. Works efficiently in a distributed data store such as Redis. Supports different refill rates and capacities for each tier.…
Read the full sample →

Rate limit designer

Input: public REST API, free and paid tiers, must survive scraping bursts

≈ 1,000
Use this
Sample outputText
Diagnosis The prompt “summarize this article” underperforms because it: Does not define the audience, purpose, or desired level of detail. Gives no criteria for identifying the article’s main point. Does not require supporting evidence, so summaries may become vague. Omits instructions for handling arguments, data, caveats, and conclusions. Specifies no output structure or length. Does not discourage generic framing, repetition, or unsupported interpretation. Rewritten Prompt — Version A: Thesis-First Summary Role…
Read the full sample →

Prompt engineer

Input: summarize this article · outputs are too generic and sometimes miss the main point

≈ 1,000
Use this
Sample outputText
Accessibility Audit Checklist — Checkout Form, Custom Dropdown, Modal Confirmation, Toast Notifications 1. Checkout Form 1.1 Errors are identified, described, and announced WCAG 2.2 criteria at risk: 3.3.1 Error Identification 3.3.3 Error Suggestion 3.3.7 Redundant Entry 4.1.3 Status Messages Keyboard test: Submit the form with required fields empty or invalid. Confirm focus moves to the first invalid field or to an error summary.…
Read the full sample →

Accessibility checklist

Input: a checkout form with a custom dropdown, modal confirmation and toast notifications

≈ 1,000
Use this
Sample outputText
Story 1: Export a report as PDF As a report user, I want to export a report as a PDF, so that I can save, print, or share a fixed-format version. Acceptance criteria Given I have permission to view a report, when I select “Export as PDF,” then the system generates a PDF containing the report’s current data and visible configuration. Given the PDF is generated successfully, when generation completes, then the file downloads with a meaningful name that includes the report name and export date.…
Read the full sample →

User story writer

Input: users want to export their reports as PDF and schedule weekly email delivery

≈ 1,000
Use this
Sample outputText
URL Shortener Architecture — Stress-Tested Revision 1. Stress-Test Findings Weak Point Category Failure or Cost Scenario Mitigation Central Snowflake-style ID Generator SPOF A single logical generator can block all automatic URL creation during failure, network partition, clock regression, or capacity exhaustion. Cross-region access also adds latency. Remove the network ID Generator. Generate random 128-bit values in each Management Service instance, encode 48–64 random bits as an eight-character Base62 code, and claim the code through a conditional…
Read the full sample →

Architecture sketch

Input: a URL shortener that must handle 10k requests/sec with analytics

≈ 2,000
Use this
Sample outputText
RECOMMENDED fix(payments): retry webhooks and correct timeout handling Transient failures caused webhook events to be dropped too early. Retries improve delivery reliability, while proper timeouts prevent hangs. fix(webhooks): add retries and resolve payment timeout bug Payment callbacks could fail permanently during temporary outages. Retry logic and corrected timeouts make processing more resilient. fix(payment-webhook): handle retries and timeouts reliably Intermittent network issues could interrupt payment notifications.…
Read the full sample →

Commit message writer

Input: added retry logic to the payment webhook handler and fixed the timeout bug

≈ 1,000
Use this
Sample outputText
Likely planner behavior and bottlenecks Postgres will normally rewrite the IN subquery as a semi-join. A plausible plan is: Scan customers to find rows where country = 'TR'. Build a hash table of matching customer IDs. Scan all 20M rows of orders. Hash-join orders against the matching customers. Sort every matched order by created_at DESC. The planner is not necessarily making a semantic mistake; without suitable indexes, this may be the cheapest available plan. The expensive parts are likely: A sequential scan of orders.…
Read the full sample →

SQL optimiser

Input: SELECT * FROM orders o WHERE o.customer_id IN (SELECT id FROM customers WHERE country = 'TR') ORDER BY o.created_at DESC · orders has 20M rows, customers 500k, Postgres 16

≈ 1,000
Use this
Sample outputText
export function add(a: number, b: number): number { return a + b; } Notes: The TypeScript version explicitly accepts and returns number. Unlike the JavaScript original, calls with strings or other non-number values are rejected at compile time. TypeScript types are erased at runtime, so JavaScript callers can still pass incompatible values unless runtime validation is added. number includes floating-point values, NaN, and Infinity. JavaScript’s usual numeric behavior still applies, including floating-point precision issues and coercion at runtime.
Read the full sample →

Code translator

Input: function add(a, b) { return a + b; }

≈ 1,000
Use this
Sample outputText
A production JSON Schema from a description: types, constraints, examples, validation notes.

JSON Schema designer

Input: an e-commerce order with items, shipping address, payment status and timestamps

≈ 1,000
Use this
Sample outputText
WebP Batch Batch-convert images to WebP from the command line with fast, consistent quality presets. <!-- Suggested badges: npm version, build status, test coverage, license, platform support --> Features Convert entire directories in one command Support for PNG, JPEG, TIFF, and other common image formats Built-in low, medium, high, and lossless quality presets Recursive directory processing Preserve source directory structure Skip unchanged or existing files Control output location, dimensions, and metadata Preview operations with dry-run mode…
Read the full sample →

README writer

Input: a CLI tool that batch-converts images to WebP with quality presets

≈ 1,000
Use this
Sample outputText
Winner: canUserAccessDocument Reading-aloud test: “Can user access document?” Clear and natural. Implies: Returns a boolean after evaluating all relevant authorization factors. Mis-implies: May sound broader than role, ownership, and sharing settings if other access rules exist. hasDocumentAccess Reading-aloud test: “User has document access.” Concise and fluent at call sites. Implies: Checks whether the current or supplied user has access.…
Read the full sample →

Naming consultant

Input: a function that checks if a user can access a document based on role, ownership and sharing settings

≈ 1,000
Use this
Video guide

Watch how it works

How AI workflows work: one form in, a finished video out

A workflow is a ready-made recipe of AI steps. You pick a template, fill in a short form, and every step runs on its own until a finished video comes out. Here we follow one bedtime story video from the form to the result.

Watch it on its own page

Chapters
  1. 0:00 What a workflow is
  2. 0:17 Pick a template
  3. 0:35 Fill in the form, see the cost
  4. 0:53 Preview cheaply, then run
  5. 1:14 Your result and run history

What happens when you press Run? The anatomy of an AI workflow

Let's open one workflow and look inside. A workflow is a graph: each box is a step, the arrows set the order, and steps that don't depend on each other run side by side. Your answers flow into the steps, some steps check their own work and run again, some wait for your approval, and every AI call is recorded with its own cost.

Watch it on its own page

Chapters
  1. 0:00 Steps, arrows and one job per step
  2. 0:26 How your answers flow
  3. 0:39 Watching it run: checks and retries
  4. 0:55 Steps that wait for your approval
  5. 1:06 Every AI call, recorded
  6. 1:18 Done, and shareable
No-code workflows

Can't find it? Build your own.

Open any template in the builder, change a step and save it as your own — or start from an empty canvas.

Start with a template Free plan · no card · the price is shown before every run Start free