Researched for Replit real features. Replace the [BRACKETED] parts with your own details, copy, and paste.
1Replit Agent full app build brief
You are Replit Agent. Build and deploy a complete working app from this description: [APP_DESCRIPTION] for [TARGET_USERS]. Stack: [FRONTEND] + [BACKEND] + Replit Database for storage. Required features: 1) [FEATURE_1] 2) [FEATURE_2] 3) [FEATURE_3] 4) [FEATURE_4]. Design: clean responsive UI with [STYLE_NOTES], working navigation, loading and error states on every screen. Data: seed the database with [SEED_DATA_DESCRIPTION]. Deployment: configure a Replit Deployment so it stays live, and verify the deployed URL serves correctly. Constraints: everything must run in this Repl with no external services except [ALLOWED_EXTERNAL_APIS]; keep secrets in Replit Secrets, never in code.
2Turn a prototype into a polished MVP
Take my existing Replit prototype ([REPL_NAME]) that currently does [CURRENT_FUNCTIONALITY] and turn it into a shippable MVP for [AUDIENCE]. Using Replit Assistant in the editor, add: user authentication (Replit Auth) with [AUTH_DETAILS], a proper database schema replacing the current [CURRENT_STORAGE], input validation and error handling on [FORMS_OR_ENDPOINTS], a landing page explaining [VALUE_PROP], and a deployment configuration so it runs 24/7 on Replit Deployments. List anything you change in a changelog. Constraints: do not break the existing working features in [WORKING_FEATURES]; keep the UI style consistent with [STYLE_REFERENCE].
3Multiplayer pair-programming session plan
We are two developers ([DEV_1_ROLE], [DEV_2_ROLE]) doing a multiplayer session in this Repl on [PROJECT_NAME], which does [PROJECT_DESCRIPTION]. Draft a 90-minute session plan: the exact files each person edits (no overlap), the shared goal [SESSION_GOAL], 3 checkpoints where we run [TEST_OR_RUN_COMMAND] together, and who resolves merge conflicts in [LIKELY_CONFLICT_AREAS]. Also write the follow-the-leader handoff notes format so the next session starts clean. Constraints: keep both developers productive with zero idle time; every task must be completable inside the Repl.
4Replit Database schema design and migration
Design the Replit Database schema for [APP_NAME], a [APP_TYPE] serving [USER_COUNT] users. Entities: [ENTITY_1] with fields [FIELDS_1]; [ENTITY_2] with fields [FIELDS_2]; [ENTITY_3] with fields [FIELDS_3]. Relationships: [RELATIONSHIP_NOTES]. Write: the schema setup code, CRUD helper functions with validation, a seed script with [SEED_VOLUME] realistic records, and a migration plan from the current [CURRENT_STORAGE] including a rollback step. Constraints: queries must stay fast at [SCALE]; no raw user input in keys; document every key pattern in a short schema README.
5One-click deployment readiness checklist
Prepare [REPL_NAME] ([APP_DESCRIPTION]) for Replit Deployments. Audit the project and produce a readiness report: 1) entry point and start command verification, 2) environment variables needed (list each with a placeholder, to be stored in Replit Secrets), 3) port and host binding check for [PORT], 4) build steps if any ([BUILD_COMMAND]), 5) files that must NOT deploy ([EXCLUDE_FILES]), 6) health-check endpoint to add at [HEALTH_PATH], 7) estimated resource needs at [TRAFFIC_LEVEL]. Then make the fixes and configure the deployment. Constraints: the app must start with a single command and survive restarts; no local-only paths.
6Bug fix via Replit Assistant with repro
Help me fix this bug in my Repl ([REPL_NAME]): [BUG_DESCRIPTION]. Reproduction: [REPRO_STEPS]. Expected: [EXPECTED_BEHAVIOR]; actual: [ACTUAL_BEHAVIOR]. Relevant files: [FILE_HINTS]. Use Replit Assistant to: 1) read the relevant code and explain the likely root cause in plain language, 2) propose the smallest fix, 3) apply it, 4) tell me exactly how to verify the fix in the Repl (what to click or run). Constraints: change as little code as possible; do not refactor unrelated code; if the cause is ambiguous, list the top 2 hypotheses with a test to distinguish them.
7Build a REST API with docs in one Repl
You are Replit Agent. Build a REST API for [DOMAIN, e.g. a bookstore inventory] with these endpoints: [ENDPOINT_LIST]. Tech: [BACKEND_FRAMEWORK] on Replit, data in Replit Database, auth via [AUTH_METHOD]. Each endpoint needs validation, proper HTTP status codes, and JSON error responses in the shape [ERROR_SHAPE]. Also generate interactive API docs at [DOCS_PATH] with a try-it-now form per endpoint, and seed [SEED_DATA]. Deploy it on Replit Deployments and give me the base URL plus a curl example for each endpoint. Constraints: follow REST conventions; every endpoint covered by the docs; no endpoint without validation.
8Classroom coding assignment setup
I teach [SUBJECT] to [STUDENT_LEVEL]. Create a Replit classroom assignment for: [ASSIGNMENT_TOPIC]. It must include: starter code in [LANGUAGE] with [STARTER_FEATURES], a clear task description with [LEARNING_OBJECTIVES], [NUM_TEST_CASES] hidden test cases that auto-grade via [TEST_RUNNER], hints that unlock in stages ([HINT_LEVELS]), and a sample solution kept separate in [SOLUTION_FILE]. Set it up for multiplayer so students can pair, and note the deployment settings if the project needs hosting. Constraints: starter code must run without errors; tests must be deterministic; difficulty appropriate for [STUDENT_LEVEL].
9Prototype a landing page and waitlist
You are Replit Agent. Build a landing page for [PRODUCT_NAME]: [ONE_LINE_PITCH] targeting [AUDIENCE]. Sections: hero with [HERO_HEADLINE], 3 benefit blocks ([BENEFIT_1], [BENEFIT_2], [BENEFIT_3]), social proof area with [PROOF_TYPE], FAQ with [FAQ_ITEMS], and an email waitlist form storing signups in Replit Database with double opt-in via [EMAIL_SERVICE]. Style: [DESIGN_STYLE], mobile-first, loads fast. Deploy it and confirm the form works end to end on the live URL. Constraints: no placeholder lorem ipsum; all copy final; form must validate emails and show success/error states.
10Scheduled job and webhook automation Repl
Build an automation Repl that does [AUTOMATION_GOAL] on the schedule [SCHEDULE, e.g. every weekday at 9am]. It should: fetch data from [DATA_SOURCE], process it per [PROCESSING_RULES], store results in Replit Database, and notify [NOTIFY_CHANNEL] with a summary formatted as [SUMMARY_FORMAT]. Add a webhook endpoint at [WEBHOOK_PATH] so [EXTERNAL_SERVICE] can trigger it on demand, secured with [WEBHOOK_SECRET_METHOD]. Include logging of each run and an admin page at [ADMIN_PATH] showing run history. Deploy with always-on settings. Constraints: idempotent runs (safe to re-run); secrets in Replit Secrets; graceful handling of source downtime.