Score AI tools for agent-readiness before integration

Clarvia AEO Check scores any MCP server, API, or CLI tool's "agent-readiness" before you add it to an agent workflow, using Clarvia's index of 15,400+ AI tools scored across four AEO dimensions: API accessibility, data structuring, agent compatibility, and trust signals. After adding the Clarvia MCP server to your config, you can score a specific tool by URL or name, returning a 0-100 score with a per-dimension breakdown, search for top-rated tools in a category, compare two tools head-to-head with a recommendation, or pull a leaderboard for a specific use case like authentication. Scores map to five bands: 90-100 "Agent Native" for tools built specifically for agent use, 70-89 "Agent Friendly," 50-69 "Agent Compatible," 30-49 "Agent Partial," and 0-29 "Not Agent Ready," meaning avoid it for agentic workflows. If a tool isn't already indexed, requesting its score by URL scans and scores it on demand rather than returning "not found."

Use it when evaluating a new MCP server before adding it to your config, comparing two tools for the same job, building an agent that selects tools dynamically, or looking for the highest-quality tool in a category. Its stated best practices: score tools before putting them into long-running agent workflows, re-check scores periodically since tools improve over time, and never skip scoring a tool just because it's well-known - popular tools can still score poorly. It also explicitly warns against using a tool scoring below 50 in a production agent pipeline without understanding what's driving the low score - use get_score_breakdown to see which dimensions are weak and judge whether they matter for your use case. It can also gate CI/CD via the clarvia-project/clarvia-action GitHub Action with a fail-under threshold.

Input: a tool's URL or name, or a category or use-case query such as "database MCP servers" or "authentication." Output: a 0-100 AEO score with a four-dimension breakdown for a single tool, ranked results for a category search, a side-by-side comparison with a recommendation for two tools, or a leaderboard for a use case.

• Clarvia MCP server (clarvia-mcp-server, run via npx)
• aeo_score and get_score_breakdown tools
• clarvia-project/clarvia-action GitHub Action for CI/CD quality gating
• Related skills: @mcp-builder for building a server that scores well, and @agent-evaluation for broader agent quality evaluation

Anyone selecting, comparing, or vetting MCP servers, APIs, or CLI tools for an agent workflow who wants an objective agent-readiness score before committing to a tool.