Let Your Agent Check if Your APIs are AI-Ready

Most engineers know the problem. You run a linter like Spectral, everything passes, and yet your AI agent still fails to use the API reliably. That is because validity and usability are not the same thing. A conformant OpenAPI document tells you nothing about whether an agent can discover, understand, and execute against it without human intervention.

That gap is measurable. And now your agent can measure it for you.

The Jentic API Scorecard, now as a native agent skill

When we launched the Jentic API AI-Readiness Scorecard CLI, the goal was to give engineering teams a concrete, trackable metric for how well their APIs serve AI agents. The CLI scores any OpenAPI document across six dimensions: structural validity, documentation quality, semantic clarity, security, and AI discoverability. It returns a single weighted grade.

The next logical step was obvious: if you are already working inside an AI coding agent, why leave it to go run a CLI manually?

You don't have to!

The Jentic API Scorecard is now available as a native agent skill for Claude Code. Install it once and your agent knows how to use the CLI: scoring OpenAPI documents, producing HTML reports, generating full diagnostics, and wiring scoring into CI, without you needing to explain the options.

See it in action

The walkthrough above shows exactly how fast this is in practice. From installing the skill through to generating a full diagnostics report, without typing a single CLI option.

Getting up and running in under two minutes

Everything you need is in the jentic/jentic-api-scorecard repository. The short version:

1. Add the marketplace and install the skill in Claude Code

/plugin marketplace add jentic/jentic-api-scorecard
/plugin install api-scorecard@jentic-api-scorecard

2. Get a free API key

Head to app.jentic.com/scorecard, create a key, and export it in your terminal:

export JENTIC_API_KEY=<your-key>

Free keys come with 100 scorings per calendar month. OpenAPI documents from the Jentic Public APIs catalogue score without a key at all.

3. Ask your agent to score an API

That is it. No CLI flags to memorise. No output format to figure out. Just ask:

Score ./openapi.yaml for AI-readiness
How AI-ready is https://petstore3.swagger.io/api/v3/openapi.json?

The skill loads automatically when your request is clearly about API scoring. You can also invoke it explicitly with /api-scorecard:jentic-api-scorecard if you want to force it into context regardless of phrasing.

What your agent can do with it

Once the skill is installed, you can ask for progressively deeper output without knowing any of the underlying CLI options. The skill teaches your agent the full interface:

• A quick headline score and grade
• A per-dimension breakdown across all six JAIRF dimensions
• Signal-level detail showing exactly which signals are dragging a dimension down
• Full diagnostics with prioritised findings by severity
• An interactive, self-contained HTML report with no external dependencies, suitable for sharing or attaching to a CI artifact

You can also enable LLM-backed analysis with --with-llm for deeper semantic evaluation: whether your API descriptions are genuinely actionable for agents, whether error responses support autonomous recovery, and more.

Why this matters

From our analysis of over 6,000 APIs, one thing is clear: most were built with a human developer in mind, and many did not adequately address even basic developer experience. There has long been a gap between what API producers deliver and what consumers actually need. That gap compounds significantly when the consumer is an agent. Agents do not ask colleagues for clarification or make informed guesses when something is ambiguous. When an API is incomplete, inconsistent, or poorly described, the results range from silent failures and unhelpful retries to agents that confidently do the wrong thing.

These are not edge cases. They are the norm across the public API landscape, and closing that gap starts with knowing where you stand.

The scorecard gives you that metric. The agent skill makes it frictionless to use inside your normal workflow, without context-switching, without manual CLI invocations, and without having to teach your team a new tool.

Score it. Fix what the score tells you. Build APIs your agents can actually use.

CLI: github.com/jentic/jentic-api-scorecard

Web UI: jentic.com/scorecard