Replit Agent

Connectors are documented as integrations Agent supports that activate automatically based on the customer's prompts, with connector management across Core, Pro and Enterprise providing centralized admin controls, scoped access and audit tracking. Dedicated warehouse connectors cover BigQuery, Databricks, Snowflake and Microsoft Fabric.

Documented use cases and build guides reach HubSpot, Slack, Gmail, calendar, Linear, GitHub, Ashby, Notion, Google Drive, Shopify, Stripe, Clerk and RevenueCat. External tools are also reachable through pre-listed MCP servers in a directory or custom MCP servers for anything unlisted.

Sourcedocs.replit.com chat/connectors, replitai/managing-connectors, connectors/warehouses, use-cases section, build/add-integrations and build/connect-via-mcp pagesread 2026-08-31

The task system documents Agent breaking work into tasks that can be reviewed, run in the background and applied back to the main version, with a dedicated guide to building several features at once through parallel tasks covering safe delegation and slicing work small.

Plan and Build modes are documented as distinct phases, Conversations and Projects as distinct surfaces with guidance on when to turn a Conversation into a Project, and Routines as scheduled recurring runs returning to a thread. Projects hold multiple artifacts, and teammates can each start their own Agent thread on a shared project. Agents and automations built on Replit run on its cloud with Slack, Telegram and scheduled triggers.

Sourcedocs.replit.com task-system, build-in-parallel, plan-vs-build-mode, conversations-and-projects, projects-and-artifacts, invite-teammates and chat/routines pagesread 2026-08-31

Replit grounds Agent through per-request context and file uploads, Custom Instructions followed across Projects, Enterprise design systems held as a DESIGN.md, Agent Skills with supporting material, and project memory retrieved when relevant.

None of these is a maintained retrieval index over the customer's documents: design systems and Skills count toward templates, project memory toward memory, Custom Instructions are guidance the product applies, and the BigQuery, Databricks, Snowflake and Microsoft Fabric warehouse connectors are the data sources for the apps Agent builds and count toward integrations. With those facts credited elsewhere, what remains is per-request context, which is Partial.

Sourcedocs.replit.com Memories page, 31 August read of chat/add-context, custom-instructions and connectors/warehouses, re-gradedread 2026-09-29

The task system page documents two vendor-shipped gates in the run path: Agent proposes tasks with detailed plans that the user accepts or revises before they start, and each finished background task runs in an isolated copy and does not change the main version until the user reviews its work log, test results and live preview and chooses "Apply changes" or "Dismiss".

That is an approval step Replit ships, not a code-host merge the customer controls. Agent creates tasks automatically when it helps, and the page does not say every main-thread edit passes the same gate. Spend controls include usage limits, per-user workspace spend limits and a per-run Routine budget.

Sourcedocs.replit.com Task system pageread 2026-09-29

An information security overview documents Replit's security practices, data protection and compliance standards, with SOC 2 Type II stated on the vendor's enterprise materials.

Documented controls include SAML single sign-on configured from workspace settings with an in-product wizard that provisions the connection from pasted identity-provider values, SCIM provisioning with synced group mapping to workspaces and roles, roles across Admin, Member, Guest and Viewer with Enterprise-only custom groups, separate account and workspace admin roles, audit logs with SIEM integration, organization-wide Enterprise privacy and source-control settings, per-user workspace spend limits, and connector management with centralized admin controls, scoped access and audit tracking. A security checklist covers securing generated applications separately.

Sourcedocs.replit.com teams/information-security, teams/saml, teams/scim, teams/groups-and-permissions, teams/account-and-workspace-admins, teams/audit-logs, teams/enterprise-privacy-settings and replitai/managing-connectors pagesread 2026-08-31

Enterprise audit logs are documented as tracking and monitoring security-relevant actions across an organization with SIEM integration. An Enterprise analytics dashboard reports activity, resource usage, published apps and costs. Connector management on Core, Pro and Enterprise is documented as providing centralized admin controls, scoped access and audit tracking. At work level the task system surfaces Agent's work as reviewable tasks that run in the background, Routines retain run history and status, and checkpoints record project state. Published apps are separately monitored after launch.

Sourcedocs.replit.com teams audit-logs, billing/teams-billing/analytics-dashboard, replitai/managing-connectors, task-system, chat/routines and build/publish-your-app pagesread 2026-08-31

The Memories page states Replit can write Memories while the user works and keep them useful over time, with user memory scoped to one builder and workspace, project memory staying with one Project and retrieved when relevant, and custom memory following its own access and visibility settings.

In Settings, Customization, Memory the user opts in, can turn Memory off, controls sharing with collaborators (off by default) and can view and update the Memory file, which is both the review surface and the purge path. In short, the agent writes its own memory, scopes it to a user or a project, and lets the user review and clear it. Custom Instructions and Skills are documented separately as instructions rather than memory.

Sourcedocs.replit.com Memories pageread 2026-09-29

Publishing is documented with autoscale, scheduled, static and reserved always-on deployment types, with release settings, live verification and post-launch monitoring. Mobile publishing is documented end to end for iOS and Android covering Apple Developer enrollment, App Store Connect upload and signing, TestFlight, Play Console upload and internal testing tracks.

Domains can be purchased and connected in-product, and development URLs cover in-progress sharing. Enterprise materials document private deployments, marketplace listings on GCP and Azure, and a Microsoft Azure partnership pairing Replit with Azure infrastructure. No customer-selectable hosting region appears in the documentation index.

Sourcedocs.replit.com replit-deployments, build/publish-your-app, mobile publishing pages, build/domain-purchasing, development-urls and enterprise materialsread 2026-08-31

Enterprise custom templates are documented as standardizing how an organization builds with Agent, alongside Enterprise design systems that configure Agent with the organization's component library, brand assets, design tokens and visual standards.

A quickstart covers building and publishing from a template, and Replit Design documents remixing a featured template into a project and a library panel offering starter templates, Mobbin-curated inspiration and reuse of the customer's own creations. Agent Skills package specialized knowledge that can be attached, installed in a project, or used to start one. The vendor's public gallery displays templates with view and use counts.

Sourcedocs.replit.com teams/custom-templates, teams/custom-design-system, build/build-and-publish-from-template, design/start-from-a-template, design/library-panel and build/use-agent-skills pages, replit.com galleryread 2026-08-31

Routines schedule recurring work from within a Conversation, using connected tools, creating projects or tasks in connected services and sending messages, with results returned to the same Conversation.

Documented limits include intervals of one hour or more, five active Routines per user on Core and ten on Pro, execution only in Power or Max mode, a per-run budget checked after each step, and a rule that a due Routine waits rather than double-queueing; Routines cannot schedule publishing. A Routines sidebar lists schedules, active status and run history with run-now and delete actions. Channels include Conversations, the prompt box, Voice Mode, the mobile app and Slack context.

Sourcedocs.replit.com chat/routines, home/routines, chat/conversations, chat/voice-mode and home/understand-your-workspace pagesread 2026-08-31

Intelligent model routing is documented as letting Replit select an appropriate model for each task, with manual selection available when the customer wants that control. A model selector page documents choosing an available model and setting its Effort, noting that manual selection depends on the customer's mode and plan.

Replit Design documents switching the model behind the design surface to explore different directions on the basis that each model brings its own sensibility and frontend strengths. Agent Modes across Free, Power and Max govern capability and spend, and Routines are documented as running only in Power or Max.

Sourcedocs.replit.com chat/auto-mode, chat/model-selector, chat/agent-modes and design/explore-with-different-models pagesread 2026-08-31

The Replit MCP Server is documented at https://mcp.replit.com/server/mcp over Streamable HTTP with OAuth through protected-resource discovery, and enumerates its tools, including create_app_from_prompt, search_apps, resolve_app_by_name, list_apps and ask_question, which create, find, inspect, update and publish Replit Apps from any MCP client.

A published endpoint, published authentication, an enumerated tool list and tools that change the platform's own apps together make Replit callable from outside, which is Full. Native connections from ChatGPT, Claude and Slack run on the same server. Outbound MCP to external servers counts toward integrations.

Sourcedocs.replit.com Replit MCP Server and Connect through MCP pagesread 2026-09-29

Agent is documented as testing the application it built in a real browser, generating a report and fixing the issues it finds before handing the work back. The artifact under test is the customer's application and the mechanism applying the criterion is Replit's, which puts it on the Full side; a run that gated only on the customer's own existing test suite would not.

Around it sit further gates in the release path: mobile testing on a real device through Expo Go and through TestFlight and Play Console internal testing tracks, a task system allowing work to be reviewed before it is applied to the main version, and checkpoints recording state for rollback.

Sourcedocs.replit.com learn/build-with-agent, task-system, build/mobile-preview-testing and learn/security-checklist pagesread 2026-08-31

A Browserbase connector is documented as running managed headless browsers for fetch, search and full session automation, for cases where Agent needs to load a real web page, run JavaScript or interact with a site that blocks plain HTTP requests. That is a third-party browser engine wired in through a connector rather than a browser Replit ships, which is Partial. Agent's self-testing exercises the application it generated in a browser and counts toward testing. No first-party browser navigation, form filling or computer-use capability appears in the documentation index.

Sourcedocs.replit.com chat/connectors and replitai/managing-connectors pages, 31 August read, re-gradedread 2026-09-29