Build a multi-tenant SaaS with an agent
The poster built a production multi-tenant SaaS in 15 days with Replit Agent, but hit issues like RLS bypasses and production pointing at the dev database, so checking its work was essential. Task: Suggested implementation: Use Replit Agent and Architect to build a multi-tenant SaaS over 15 days. The workflow involves writing a spec, converting it to markdown, having Replit propose a plan, sanity-checking with Perplexity or Codex, and executing via checklist. The human role is critical for constant verification and reviewing agent outputs.; tools: Replit Agent. Human role: The original post does not say how a person stayed involved.; limitations: Prerequisites include strong supervision capabilities, as the agent cannot read production logs. The main constraint is the high cost of constant verification. Integration uncertainty exists regarding security configurations like RLS and OAuth. Human approval is required for all critical security and environment settings..
Tools
Published on Agent Work Atlas:
Overview
The poster built a production multi-tenant SaaS in 15 days with Replit Agent, but hit issues like RLS bypasses and production pointing at the dev database, so checking its work was essential.
Workflow
Suggested implementation: Use Replit Agent and Architect to build a multi-tenant SaaS over 15 days. The workflow involves writing a spec, converting it to markdown, having Replit propose a plan, sanity-checking with Perplexity or Codex, and executing via checklist. The human role is critical for constant verification and reviewing agent outputs.
Reported result
The SaaS was built, but the speed was misleading due to unexpected failures in every layer. Specific issues included RLS bypasses, double-encrypted OAuth tokens, reverted environment variables, and production pointing at the dev database. Actual outcomes indicate high supervision overhead rather than seamless automation.
Human role
The original post does not say how a person stayed involved.
Limitations
Prerequisites include strong supervision capabilities, as the agent cannot read production logs. The main constraint is the high cost of constant verification. Integration uncertainty exists regarding security configurations like RLS and OAuth. Human approval is required for all critical security and environment settings.
Try it
Act as a senior developer using Replit Agent to build a small multi-tenant feature. First, write a detailed spec in markdown. Ask Replit to propose a plan, then verify this plan with Perplexity for security gaps. Execute the plan step-by-step. After each major step, manually verify Row Level Security (RLS) policies and environment variables. Do not deploy to production until you have confirmed that production does not point to the dev database. Measure the time spent on manual verification versus agent generation. If verification time exceeds 50% of total time, pause and refine the spec. Output a report detailing any security bypasses found.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.