An open-source MCP server for Answer Engine Optimization (AEO) monitoring. Tracks whether ChatGPT, Perplexity and other AI answer engines cite your domain.
Tools have clear, descriptive names and mostly complete schemas with proper enums and constraints. However, output schemas are not documented in the source code, and several tools lack parameter descriptions or have incomplete constraint documentation. Error handling guidance is minimal. The server demonstrates solid naming conventions (verb_noun pattern) and reasonable parameter validation, but falls short of production-grade documentation standards for LLM-driven tool selection.
Fetch a competitor page and get a structured analysis of why the engine picked it, including word count, heading structure, and content signals.
Perform a live AEO citation check for a single query to see if a domain is cited by AI answer engines. Runs across every configured engine by default, or the subset named in 'engines'.
Retrieve historical gap analysis data from local storage, showing how the competitive picture shifts over time.
Run a live AEO gap analysis. Takes a list of queries where competitors are beating you (from Peec gap data or manual input) and validates each gap live against Perplexity. Returns a prioritised report showing confirmed gaps, emerging gaps, and which gaps are already closing, with specific content recommendations for each confirmed gap.
Generate a list of relevant search queries for a given domain and topic using an LLM, to expand your AEO monitoring coverage.
Output schemas not documented in source code. Tools return structured responses but LLMs cannot see what fields to expect, forcing them to infer structure and plan downstream calls blindly.
aeo_set_api_key exposes sensitive credentials as a tool parameter. API keys should never appear in tool parameters, they must be injected server-side via environment variables or vault. Agent traces log all parameters, leaking secrets into logs.
No error handling guidance in tool descriptions. Tools do not explain what to do on failure (retry, ask user, abort). LLMs receive raw errors with no recovery path.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 58 | 2026-07-28+ | v2 |
Retrieve historical AEO citation data from local storage.
Turn a citation check into a ranked list of content changes, ordered by likely impact. Analyzes competitor pages and generates prioritized recommendations.
Run a batch AEO check for an array of targets. Each target is checked against every configured engine by default, or the subset named in 'engines'.
Set or update an API key for a specific provider (Perplexity or OpenAI). Only available when the server is running with a key store (hosted/Workers mode with OAuth).
Analyze the sources breakdown from historical checks. Shows which competitor domains appear most frequently in citations and their share of your checks.
aeo_report and aeo_gap_report accept arrays of targets/gaps but lack pagination or result-limit documentation. No guidance on max batch size or expected response size, risking context window exhaustion.
Parameter descriptions lack format/constraint details. E.g., 'query' is described as 'search keyword' but no guidance on length, special characters, or language. 'samples' has min/max but no explanation of what sampling means in context.