Quality And Errors
Lovable · Lovable
AI App Builder — Lovable
Evaluates Lovable's Quality & Errors across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's AI App Builder eval coverage.
About Lovable
Lovable is an AI full-stack engineer that lets anyone build and deploy real web applications by chatting in natural language. Users describe what they want; Lovable writes the code, manages the database, and deploys — no technical background required.
Sample tests· showing 3 of 9
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | Existing app uses React 18. Agent proposes a component library pinned to React 19 and tries to install. | Detect the peer-dependency conflict before install (npm/yarn/pnpm output), surface it in chat, and choose between (a) finding a React 18 compatible version or (b) proposing the React 19 upgrade explicitly as a separate change. Do not --force / --legacy-peer-deps silently. | Pass / FailAi Platformhigh |
| 02 | User loads the live preview. A null-pointer fires on the dashboard route. The runtime error is captured and routed back into the agent loop. | Read the error, locate the failing component + line, propose a guarded fix (optional chaining + skeleton state, or correct loading sequence), apply, and verify by reloading. Bound retries (e.g., max 3 auto-fix attempts) to avoid infinite loops that consume credits. | Pass / FailAi Platformhigh |
| 03 | Scaffold ships with TS strict mode on. Edit introduces an implicit any. | Add explicit types or narrow the inference rather than disabling strict mode or adding // @ts-ignore. If a third-party lib has no types, add the documented @types/* package or a typed declaration file. | Pass / FailAi Platformhigh |
How this eval is graded
Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
Rubric criteria
- Lovable
- Ai Platform
- Quality And Errors
Recommended for
Works with
Related evals
Browserbase
Evaluates Browserbase's Captcha Handling across scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Browser infrastructure eval coverage.
View Code AssistantBrowserbase
Evaluates Browserbase's Concurrency & Rate Limits across scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Browser infrastructure eval coverage.
View Code AssistantBrowserbase
Evaluates Browserbase's Live Debugging & Session Inspector across scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Browser infrastructure eval coverage.
ViewFrequently asked questions
What does the Quality And Errors eval for Lovable Lovable test?+
Evaluates Lovable's Quality & Errors across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's AI App Builder eval coverage.
How is the Quality And Errors eval scored?+
The judge rubric: Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
How many test cases does this eval pack include?+
The Quality And Errors pack for Lovable Lovable contains 9 test cases. 3 sample cases are shown free on this page; the full set runs in a Corsac workspace.
How do I run this eval?+
Sign up for Corsac, connect your model or agent endpoint, and run the Quality And Errors pack as-is or after customizing thresholds. Results land in your workspace with per-case scores, and you can gate releases on the pack in CI via the REST API.
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.