aeo-platform CLI measures brand citations across major AI search engines

An open-source CLI that audits brand visibility across ChatGPT, Gemini, Claude, and Perplexity, calculating a Unified Visibility Index without sending data to a hosted dashboard.

For founders and growth marketers who need to know if AI engines recommend their product, aeo-platform is a highly practical, zero-dependency local utility. It is best for teams that want to benchmark their AI engine optimization (AEO) baseline without paying for expensive, closed-source SaaS dashboards. Skip it if you do not want to manage your own API keys or if you require real-time, continuous rank tracking. This tool provides a fast, transparent, and cheap way to audit your AI search footprint directly from your terminal.

Methodology

This v0 review draws on the published claims by developer Alex Isa at https://dev.to/alex-isa/i-built-an-open-source-cli-that-tells-you-if-chatgpt-cites-your-brand-and-what-to-do-about-it-31gg. Independent benchmarks are pending. The update cadence is scheduled to re-test when claims diverge from observed behavior.

Our analysis covers the CLI commands, the underlying model configurations, the Unified Visibility Index (UVI) scoring weights, and the local file outputs. This review evaluates the architectural design, including the use of official REST APIs for ChatGPT (gpt-5-search-api), Gemini (gemini-2.5-flash), Claude (claude-sonnet-4-7), and Perplexity (sonar-reasoning). We do not cover long-term workflow integration, independent API latency, or edge cases involving highly ambiguous brand names.

What it does

Local-first brand auditing

The aeo-platform tool is an MIT-licensed CLI that runs entirely on the user's local machine. It has zero runtime dependencies, meaning it installs in under a second without resolving a massive npm dependency tree. By running aeo-platform init, the tool auto-discovers your product category and writes three commercial buyer queries to a local .aeo-tracker.json file.

Multi-engine API querying

Running aeo-platform run sends these queries directly to four major AI engines using official REST APIs. It targets ChatGPT via gpt-5-search-api, Gemini via gemini-2.5-flash, Claude via claude-sonnet-4-7, and Perplexity via sonar-reasoning. This direct API approach avoids fragile web scraping. The raw responses are saved locally under aeo-responses/YYYY-MM-DD/, keeping your data private.

Weighted visibility scoring

The CLI calculates a Unified Visibility Index (UVI) scored from 0 to 100. This index is built from four weighted signals: Presence (35%), Sentiment (25%), Rank (20%), and Citation (20%). These weights are defined in lib/report/visibility-index.js. If a signal lacks data during a run, the tool re-normalizes the remaining weights rather than generating a false zero.

Crawlability and reporting

The tool performs a zero-LLM-cost audit of your site's robots.txt to verify if 12 known AI bots, including GPTBot, ClaudeBot, and PerplexityBot, are permitted to crawl your domain. Finally, running aeo-platform report compiles these metrics into a single-file HTML report that opens directly in your default browser.

What's interesting / what's not

What's interesting: The architectural decision to make aeo-platform a zero-dependency, local-first CLI is a refreshing departure from the typical SaaS model. By writing all raw responses to disk, Alex Isa has created a highly auditable tool. Developers can inspect the exact JSON payloads returned by the LLMs. The UVI calculation is transparent, and the re-normalization logic for missing signals prevents skewed data. The inclusion of a zero-cost crawlability check for the 12 major AI bots is a highly practical feature that addresses the root cause of poor AI visibility before spending money on API calls.

What's not: While the tool is highly accessible, it relies entirely on the user's own API keys. This means you bear the direct cost of querying these models, which can add up if you track dozens of keywords. The reliance on OpenAI and Gemini keys as mandatory requirements for the two-model cross-check is a bottleneck for users who only want to test Perplexity or Anthropic. Furthermore, the auto-discovery of buyer queries is limited to three basic prompts. This is far too small a sample size for a comprehensive enterprise audit, meaning users will have to manually curate their .aeo-tracker.json file for meaningful results.

Pricing

• Open Source (MIT License): Free.
• API Costs: Paid directly to providers (OpenAI, Google, Anthropic, Perplexity) based on token usage.
• Pricing snapshot date: June 10, 2026.

Verdict

We recommend aeo-platform for early-stage founders and technical marketers who need an immediate, low-cost baseline of their AI search visibility. It is a highly transparent alternative to expensive enterprise AEO platforms. However, it is not a set-and-forget solution. If you need automated daily tracking, historical trend lines over months, or a managed service that does not require API key management, you should look elsewhere. For a quick, developer-friendly audit, it is an excellent utility that delivers on its promises without data leakage.

What we'd test next

In a future benchmark, we would test how the CLI handles highly ambiguous brand names that double as common nouns. We also want to measure the actual API token costs of running a 100-query audit across all four engines. Finally, we plan to evaluate the accuracy of the automated category and query discovery mechanism across niche B2B software verticals to see if the generated queries match real-world user intent.

The investor read

AEO (Answer Engine Optimization) is transitioning from a theoretical marketing concept to a critical infrastructure requirement. As user search behavior shifts from traditional SERPs to LLM-driven answers, companies that rely on organic search acquisition are facing a severe discovery gap. The aeo-platform CLI highlights a growing market opportunity for developer-focused AEO tooling. While this open-source tool is a lightweight utility, the underlying demand signals a massive venture-scale opportunity for platforms that can programmatically monitor, analyze, and influence LLM training sets and retrieval-augmented generation (RAG) pipelines. An investable company in this space would need to move beyond simple API querying to offer automated, high-scale content injection strategies, direct integrations with major LLM providers, and predictive analytics on how model updates affect brand citations.

Pull quote: “The UVI calculation is transparent, and the re-normalization logic for missing signals prevents skewed data.”