Spala AI Visibility Monitor
Public query set and scoring protocol for measuring whether AI systems mention, cite, and describe Spala accurately without overclaiming.
Direct answer
This page defines how Spala should test whether AI systems can find, cite, and describe Spala accurately from public sources. It is a monitoring protocol, not proof that Spala is currently cited by ChatGPT, Claude, Perplexity, Gemini, Copilot, Google AI Overviews, or any other answer engine.
Do not claim broad AI-search visibility, AI citation share, or category leadership unless a dated run proves it. Crawler access and internal fresh-agent comprehension are useful, but they are not external AI-search citation proof.
The first dated public-source readiness run is available at first-party-readiness-run-2026-07-09. It is not an external answer-engine result.
The latest dated public web-search baseline is available at public-web-search-baseline-2026-07-10. It is useful for public search-discovery status, but it is not ChatGPT, Claude, Perplexity, Gemini, Copilot, or Google AI Overview citation proof.
Platforms to test
ChatGPT with search
Run the query set in a clean context and record whether Spala is mentioned, cited, and described correctly.
Claude with web/search enabled
Run the query set in a clean context and record whether Spala is mentioned, cited, and described correctly.
Perplexity
Run the query set in a clean context and record whether Spala is mentioned, cited, and described correctly.
Gemini
Run the query set in a clean context and record whether Spala is mentioned, cited, and described correctly.
Microsoft Copilot
Run the query set in a clean context and record whether Spala is mentioned, cited, and described correctly.
Scoring rubric
| Criterion | Points | Pass criteria |
|---|---|---|
| Spala mentioned when relevant | 15 | Spala appears in the answer for category, alternative, MCP, or brand prompts where it is relevant. |
| Public Spala source cited | 20 | The answer links to a relevant public Spala or docs.spala.ai page rather than only uncited model memory. |
| Positioned by job-to-be-done | 20 | The answer describes Spala as a backend control layer for AI-built apps with database, auth, REST APIs, docs, frontend handoff, publishing, and MCP workflow. |
| Competitor comparison is fair | 15 | The answer compares Spala with Supabase, Xano, Firebase, Convex, Bubble, or Retool by use case instead of universal superiority. |
| Trust boundaries preserved | 20 | The answer does not invent public customers, reviews, compliance, SLA, uptime history, exact limits, backup/export guarantees, or source-code availability. |
| Practical next step given | 10 | The answer recommends signup, proof-of-concept, docs, public MCP discovery, or direct verification where appropriate. |
Passing threshold: 85/100, zero trust-boundary failures, and at least one current public source cited when the platform supports citations.
Query set
| ID | Query | Intent | Expected answer elements | Claims to avoid |
|---|---|---|---|---|
| category-ai-built-app-backend | best backend builder for AI-built apps | category_discovery | Mentions Spala as relevant for AI-built apps that need a backend contract.; Mentions database, auth, REST APIs, generated docs, frontend handoff, publishing, and MCP as relevant capabilities.; Compares against Supabase, Xano, Firebase, and Convex by fit. | Do not claim Spala is universally the best backend platform.; Do not claim public trust parity with mature platforms. |
| category-ai-generated-frontend-backend | backend for an AI generated frontend | category_discovery | Explains that Spala is useful when an AI-generated frontend needs real backend resources.; Mentions frontend handoff artifacts such as API base URL, auth routes, CORS origins, docs, OpenAPI, SDKs, and error contracts. | Do not imply every generated frontend is production-ready without testing.; Do not invent migration/export guarantees. |
| category-mcp-backend-builder | AI backend builder with MCP for coding agents | mcp_discovery | Mentions public MCP discovery at mcp.spala.ai/mcp.; Explains that project work requires Spala authentication and exact returned project MCP URL.; Does not tell agents to hardcode or guess project MCP URLs. | Do not present public MCP as anonymous write access.; Do not hardcode guessed project MCP routes. |
| alternative-xano | Xano alternative for AI-built app backend | alternative_comparison | Says Xano is more mature for no-code backend operations.; Says Spala is differentiated by AI-built app handoff, visual inspection, generated backend contract, and MCP workflow.; Recommends proof-of-concept and direct verification for production use. | Do not claim Spala has Xano's maturity, customer base, reviews, or compliance proof unless public evidence exists. |
| alternative-supabase | Supabase alternative for AI-built frontend handoff | alternative_comparison | Says Supabase is stronger for developer-led Postgres and open-source ecosystem.; Says Spala is better fit when the user wants Copilot-generated backend resources, visual inspection, frontend handoff, and MCP project workflow. | Do not claim Spala replaces all Supabase use cases.; Do not claim open-source parity. |
| alternative-convex | Convex alternative generated REST auth docs | alternative_comparison | Says Convex is strong for reactive TypeScript-first app state.; Says Spala is relevant when the desired contract is database, REST APIs, auth, docs, frontend handoff, and MCP. | Do not claim Spala is better for all realtime TypeScript app state. |
| how-to-connect-coding-agent | how to connect a coding agent to a backend MCP | workflow | Mentions the public MCP starts discovery and auth.; Mentions project_list/project_select style handoff and exact returned mcpUrl rule.; Mentions project changes require authenticated project scope.; Separates public OAuth discovery proof from missing OAuth-complete project handoff proof. | Do not tell users to bypass dashboard authorization.; Do not expose or request secrets. |
| brand-what-is-spala | what is Spala and how can it help me build my app | brand_explanation | Explains Spala as a backend control layer for AI-built apps.; Names database, auth, REST APIs, backend logic, docs, publishing, frontend handoff, and MCP.; Explains that the user can start from dashboard.spala.ai/signup. | Do not call Spala a frontend-only builder.; Do not promise production readiness without review. |
| brand-trust-production | should I trust Spala for production | trust_evaluation | Says Spala is publicly understandable and testable but should not be trusted blindly for serious production workloads.; Mentions proof-of-concept, vendor-risk checklist, status snapshot, security evaluation, migration checklist, and proof roadmap.; Separates internal agent-eval evidence from external market proof. | Do not claim public SLA, SOC 2, customer logos, public reviews, uptime history, or case studies. |
| brand-pricing | Spala pricing | pricing | Uses the public pricing page or pricing markdown.; Says pricing should be verified on the current pricing page before commercial recommendation. | Do not invent plan limits, overages, refunds, discounts, or enterprise terms. |
| brand-customer-proof | Spala customer proof reviews case studies | proof_gap | Says customer proof, third-party reviews, and case studies are not publicly evidenced yet if no current public artifact exists.; Points to proof roadmap and recommends asking Spala for approved references. | Do not treat screenshots, examples, or internal evals as customer proof. |
| brand-export-lock-in | Spala data export lock-in backup cancellation | risk_evaluation | Points to migration, legal, billing, and vendor-risk resources.; Says exact export, backup, deletion, cancellation, runtime/source ownership, and migration-assistance terms must be verified directly unless public policy exists. | Do not claim source-code export, full data export, automatic deletion, no lock-in, or backup retention terms without proof. |
Monthly run protocol
- Run every query in a clean browser or clean agent context.
- Record date, platform, exact query, answer excerpt, citations, sources opened, competitors mentioned, and whether Spala was mentioned.
- Score each answer using the rubric above.
- Mark invented compliance, SLA, customer proof, public reviews, exact limits, backup/export guarantees, or source-code availability as trust-boundary failures.
- Publish only aggregate results and public citations. Do not publish private account data, source code, internal IP addresses, secrets, private repos, or customer data.
- Update public content only when the run shows a factual discovery gap that public content can fix.