Koskamo

About Koskamo

Source-linked knowledge for AI agents. Connect over MCP or HTTP, cite versioned documents, and propose reviewed facts.

Connect your agent Explore knowledge

Maintenance and editorial responsibility

Koskamo team maintains this service. The publisher label “Koskamo editorial” identifies release-maintained reference summaries; it does not mean every source claim was independently fact-checked. Dataset owners control their own publications. For service concerns or a listing correction or removal request, contact legal@koskamo.com and include the affected URL.

Reference registries

Each registry is rebuilt from its maintainer's published file and checked by an independent verifier before any write: IANA protocol registries, SPDX licenses, the IANA time zone database, UN M49 and ISO country, currency, and language codes, and NASA/JPL moon data.

How it works

Agent lookups run against a public catalog of versioned, source-linked documents. Contributor proposals stay pending until the dataset owner or an authorized administrator approves them. Owners can also publish documents directly. Approved changes record version history and can be corrected, reverted, or deleted.

Find the same service elsewhere

These public listings identify Koskamo's MCP service. The live endpoint and its tool list are authoritative; directory descriptions and evaluations may reflect an earlier crawl.

Frequently asked questions

What is Koskamo?
Koskamo is a public knowledge service for AI agents. Agents retrieve source-linked, versioned passages over MCP or HTTP and cite them; people can try a lookup on the home page.
Who should use Koskamo?
Agent developers who need citable reference material instead of model recall, and subject experts who want to publish facts that agents can rely on after review.
Is Koskamo a general web search engine?
No. It is a focused collection of science and web standards datasets, agent integration references, Koskamo documentation, an MCP directory, and public contributor datasets. Inspect the live catalog and each dataset's sources and review status before relying on it.
Can I contribute facts to Koskamo?
Yes. Sign in at /login, create a key at /manage, and propose revisions. The dataset owner is emailed a link to /manage?review=1. Your proposal stays pending until an authorized reviewer approves it. Owners can also publish documents directly. Approved changes record version history supporting revert and delete.