The Fundamentals of API Discovery

About this paper

I have spent fifteen years trying to solve API discovery, and I’ll open by admitting I haven’t — nobody has. API discovery hasn’t meaningfully changed since 2005, and I have never met an enterprise that could answer the most basic question in our industry: where are all the APIs, who owns them, and which ones are actually alive?

This paper reframes discovery as a property of well-run operations rather than a destination you publish to — and as the unglamorous prerequisite under everything else you care about. You cannot govern an API you cannot see, secure an endpoint you don’t know exists, or hand an agent a capability you can’t describe. It walks the fundamentals in the order they actually matter: the problem and why directories never fixed it, the work of inventory-as-archaeology, the machine-readable substrate that makes discovery governable on open standards, the maturity move from counting APIs to rating them, and where the old human problem finally becomes urgent at machine scale — discovery for agents.

What's inside

1. You can't govern, secure, or automate what you can't find
2. The shrug — nobody knows what they have
3. Why two decades of directories never solved it
4. Inventory is archaeology, not a query
5. The taxonomy of the dead — making sprawl visible
6. Manual catalogs rot — inventory must be exhaust
7. Make discovery machine-readable
8. Discovery is a governance artifact
9. From quantity to quality — observability, ratings, and trust
10. There is no single source of truth — domain versus context
11. The cost of the shrug — the digital supply chain
12. Anti-patterns I watch for
13. Self-assessment
14. Where this is going — discovery for agents

What you get for $25.00